You are here

NEScaf: A Switched Capacitive Audio Filter

Contributed by Dave Siegrist NT1U

The NEScaf is a switched capacitive audio filter. It is designed to be simple to build and use.

NEScaf PC boardIt is more flexible than previous filters for QRP rigs. It has two panel-mounted controls — a single center-detent potentiometer that sets the center frequency of the filter's bandpass (default value settable between ~415Hz and ~1kHz), and a dual-ganged potentiometer that provides a continuously variable bandwidth control (from ~60Hz to ~1500Hz). This ability to smoothly vary the bandwidth is very useful.

The NEScaf is designed to be driven with a phones or speaker level signal. No hacking into your rig — just connect it to the headphone or speaker output of the radio. This makes it usable with several radios. The board is small enough, at about 2 inches square, to build inside many radios however, if you so choose.

This kit is suitable for all experience levels. It uses all thru-hole components, and the construction and adjustment is very straight-forward. It is reasonably priced and will be a nice addition to the QRPer's arsenal of tools.

The NEScaf is only $31 postpaid in the U.S. For your convenience, we offer a connector set for just $4. Visit the NEQRP Club Store to purchase the NEScaf.

Theory

The integrated circuit at the heart of the NESCaf is made up of two CMOS active filters. These filters are extremely configurable (low pass, band pass, notch etc). We have chosen to set up both filters as Butterworth band pass filters and to cascade the filters. Butterworth filters have the characteristic of constant amplitude in the band pass region, while the cutoff knee is not be as sharp as if the filter were configured as a Chebychev. We considered this an acceptable tradeoff, wanting constant volume out regardless of the bandwidth or center frequency setting of the filter. There are two on-board trimmer pots. One is used to "calibrate" the center frequency pot. This allows you to adjust the frequency at which the center detent occurs. If you are using a rig with a transmit offset and sidetone of say, 700 Hz, you can use this trimmer to make that the center-detent frequency. The other on-board trimmer adjusts the audio level into the filters output amp. Using this pot, you can set the overall gain of the filter. This can be used to set the filter for unity gain, if desired. This way, the filter could be switched in and out, and still maintain a comparable volume level in the headphones. The NEScaf has no panel mounted volume control of its own. It is designed so that you use the volume or RF gain control (if present) on the rig it is being used with.

Fine Print

Please Note: A case and audio/power connectors are NOT supplied. Since each builder will have his or her own personal favorite styles for these, we are not including these parts. All board-mounted parts, plus the two panel-mounted potentiometers, a power/bypass switch, and a power-indicator LED are included in the kit. You supply your own case. You can supply your own connectors for audio in and out and power, or you can order a connector kit from us with your NEScaf filter.

The NEScaf was originally kitted with the MF10. That was replaced by the LMF100, a drop-in replacement for the MF10. These were superseded by the LTC1060 in spring 2010.

Comments

I saw two new NEScafs at our local Saturday morning ham breakfast today. They were built by Gene (AB1FN) and his son Vince (KB1JHG) of Charlestown, NH. Both got the kits at Boxboro last weekend. Gene won his kit at the QRP forum.

Both looked great. Gene built his into a nice black enclosure with beautiful panel labeling he did. He added a "VU" signal level meter to the front panel. It looked really slick.

Vince had built a 40M RockMite into an Altoids tin and wanted to add the filter. He had previously removed all the paint from the RockMite's tin. He did the same to a second tin, and built the NEScaf into that. he then soldered the two tins together to make a nice compact package.

Both gents said that they really liked the performance of the filter. I hope to have pictures to post of these sometime soon. Those who have built a NEScaf and would like to share a photo can post it to the NEScaf Gallery page.

OK one more up and running on the Left Coast this time. Set it up for my NorCal 40A and centered the pass band on the side tone. Besides the noise reduction, it really does indicate "zero beat" with the filter set with a very narrow pass band.

Neat project, thanks for doing this guys!!

73

Doug W7RDP

2/2/07 OK I installed the LMF100 and the CMOS 555, not sure what the CMOS 555 did but the LMF100 is much more quiet than the LM10, very big decrease in noise level. I will have to see if MAXIM will send me some samples of their version to compare with the LMF100.

Still a very cool project, you guys just need to stock up on some XL t-shirts ;-)

73, Doug W7RDP

Steve KA4NRF, Came back from Boxboro with the kit and before I could build it received the 2 LCD freq display kits so I was swamped. So between the Kids and work finally got around to building all 3 kits and the NEScaf is awesome,share it with my FT817 and SW40+. Worth its weight in gold.

hi guys,

I really liked the New England QRP club's switched capacitor audio filter kit,

but i thought it could be better, a little quieter. at first i thought the lm386 was adding some hiss. instead of using the lm386 i used a tda2003. used the circuit from page 27 of the sprat passed out at lobstercon, titled "High Performance Headphone Amplifier". i think most of the noise thats added is coming from the scaf itself and not just the 386. if the band is noisey its not objectionable but if the band is very quiet and especially if the signal is not very strong then there is room for improvement. regardless I still find it very useful for figuring out where zero beat is, and i have been amazed at how well i can sometimes isolate a signal from nearby interference.

if you used a 20 pin socket (radio shack 276-1991) then replacing the scaf chip is very easy. National made the MF10 supplied with the kit. if you go to their site they recomend an LMF100 as an upgrade. this is a 'pin compatible' drop in replacement. couldn't find it at mouser but they did have them at digikey for $6.09 for one (in addition to tax and shipping, digikey adds a $5 charge if your order total is less than $25). the other alternative drop in replacement is a MF10 made by maxim.
http://www.maxim-ic.com/parts.cfm?p=MF10
couldn't find them for sale anywhere but if you sign up with maxim you can request a sample, the MF10CN and MF10BN are 20 pin dips, not sure about the others or what the cn & bn mean.

i have tried both the LMF100 and the maxim chips and they are quieter than the National MF10. not sure which is the best as both the replacements seem about the same to me but enough different from the stock National MF10 that i think you will notice it and be pleased with the improvement.

72
jerry terres aa1of
franconia nh

(with copy to QRP-L)

I have built the NEScaf into 3 rigs, and find it terrific. I am ordering one more to set up separately. VERY useful kit!

John-W2AGN

The minute I put power to my newly built SCAF, it came to life and worked! I guess you could say it was CLEAN!

Dennis--K1LGQ
Brookline, NH

Hi Jerry,

as I checked with Maxim (even directly with their manager) they are not able to produce the MF10 -> they are not sending the samples MF10BN/CN as You mentioned.
Can You specify where You got them please?

73, Petr OK1RP

(continued)
under "annual usage" i estimated " click "continue"
click "Register if you are a new user"
register for membership by filling out the contact information and click "submit" then check your email.
To my ears the Maxim brand MF10 sounded as good as the MF100. If the kind folks at Maxim decide to send you a sample then you can decide for yourself.
72
jerry terres aa1of
franconia nh

Jerry,
thank You for reply. I know how to order Maxim samples and I did exactly as You mentioned but...instead of the shipping I got message from manager of Maxim that they are not able to produce that product and that is not possible to ship any samples. Even for my second request I got reply that there is no way.

73, Petr OK1RP

I recently received my NEScaf kit, and the .33uF capacitor is missing. I was unable to locate a like-value replacement at Radio Shack. Is there a substitution value that is suitable for this component; or can I get a replacement from NEQRP?

Thanks,
Roger Lange, KC3DI

Hi Roger.

Sorry for the delay in replying. I have forwarded your message on to Dave, NT1U. He should be able to get you a replacement cap shortly.
73,
Bruce N1RX

Nice Kit! I have it hooked to my Rock Mite 40m. The only snag I have run into is setting it up, I find the instructions here Very confusing.

Page 49(!) says to hook the right tab, front view, to the "correct" circuit board location. The schematic shows the wiper hooked to one of the other tabs, which one, who knows! So it is anyones guess which way to hook it up.

Page 59 says to adjust the Bandwidth (R7?) to it's widest position, not sure which way, I adjusted it for loudest sidetone.

Then it says to narrow the bandwidth and adjust the center freq. to keep it audible. It then says to set the BW to its narrowest setting and peak the audio with the Center Freq control. What is the point? Why not set the BW to minimum and peak with the CF. What are we doing here?

What about R9, the pot on the PCB going to the 555? No mention of it's setting that I can find.....

Thanks for the help

Rich

Hi Rich.

I don't see any other replies yet, so here goes.

First off, the manual needs some updating/correction. When reading the manual, "page49" should be read as "4 of 9". Editing oversight. On to your specific questions...

R7 is the front-panel mounted, dual section potentiometer. It will control the bandwidth of the filter. Each section (front and rear) are wired to a separate set of holes on the PC board. For each section, the "wiper" or middle lead, goes to either R7a_W or R7b_W. doesn't matter which, as long as you wire the corresponding "end" connection for each section to it's corresponding pad.

So, if you call the front section "a", then the wiper of the front-most pot section will go to point R7a_W. One of the end connections will go to R7a. Which "end" of each pot section you use depends upon your personal preference for how you want this control to behave. The filter is at it's narrowest (minimum bandwidth) when the pot is turned towards it's minimum resistance end.

If you want the filter to get narrower as you turn the pot clockwise, then use the pot connection that the wiper moves toward when you turn the pot clockwise. If you want the filter to get wider as you turn the knob clockwise, then use the other end. The important thing is to wire both pot sections the same. You can always swap the leads over to the other side of the pot if your preference changes.

Re page 59 (5/9)bandwidth pot initial setting: The intent here is to initially set the bandwidth pot (R7) to it's widest setting. Which way to turn it will be determined by how you wired the pot in the step above.

Thoughout the procedure that follows, the front panel frequency pot (R10) should be left in it's center-detent position. The frequency pot pot being adjusted in this process should be R9, the board mounted trimmer for the 555 clock generator. The goal through this procedure is to end up with the center detent spot on the front panel pot to be set at your sidetone (transmit offset) of choice.

I hope this helps to clarify things.

73,
Bruce N1RX

Thank you Bruce, that helps a lot. I think the key missing bit was any mention of R9.

I will try it again. Other than the setup instructions the filter kit is very nicely done and a real value.

73

Rich

They may or may not be willing to send you a sample, its up to them. the "+" designation at the end of the part number, for Maxim, means lead free. the BN version is a DIP (dual inline pin vs surface mount type).to request a sample try using this link: http://www.maxim-ic.com/parts.cfm?p=MF10
next to "MF10BN+", click on "sample"
on the resulting page click on "sample now"
up near the top of the same page click on "Sample Cart"
put what you want under "customer part or reference number#"
under "sample quantity" change to "1" if you want (2 is ok)
under "annual usage" i estimated "
(contiued)

Jerry,
thank You for help -> but again...there is copy paste of the reply from Maxim:

Petr,

We would like to sample you this product, but we are no longer manufacturing this part. We do have many filters, perhaps you could browse our webpage. Maxim-ic.com and find an alternative. From what I know, the MF10 series has been killed for more than 6 months.

Thank you,

Kyoung Kim
Maxim Integrated Products
Inside Sales Associate
Phone: (408)331-4222
eMail: Dongjoo_Kim@maximhq.com

Just finished building the NESCAF and wow, it works great. Received it in 2004 I think and finally got around to it. It works really well. I hooked it to my MicroR2 receiver that was just completed a couple of days ago. Makes a great combination.

Just ordered the M100 chip. Thanks for a great kit!!

John Arnold WA6YSY
QTH: Ocean Shores, WA abt 75 miles West of Olympia

John wb1hbe

Think I will give this kit a try Dennis.

I just completed my NESCAF and I've tested it out. I built it to use with my backup rig, an Alinco DX-77T. I have to say that I am completely impressed. Not only has it worked well for me on CW, but I cranked open the bandwidth, moved up the center frequency, and listened to some SSB using it.

I put my unit in a Bud cast aluminum enclosure painted gray to match a passive SSB LP filter I built previously. With both of these in line the band noise is cut down considerably.

Excellent job to all who put this one together.

Andy N1KSN
NEQRP #176

I made a soldering error while building the kit. Has anyone had luck with desoldering from the board?

Hi Bill.

There is nothing special about the PC board for the NEScaf that makes it particulary difficult to remove a component. I'm not sure what you mean by "has anyone had luck..."

Use enough heat, and your favorite solder removal method (i.e. solder-wick desoldering braid, or solder sucker etc.) I have found that in general, it can be helpful to add a little more solder to a joint when attempting to remove a component. This allows the newly-flowed solder to help conduct heat into the joint, and introduces a little more flux for better flow.

If there is a specific question about parts removal from a PC board, then please repost back with some specifics. I am sure you will be able to get your questions answered.

GL,
Bruce N1RX

Can I just bypass the Vreg and power it from a PP3 battery?

73 cheers
Nick m0NjP
SW England (old)

Hi Nick. I assume a "PP3" battery is a 9 volt type? If so, then the answer is yes. The Nescaf can be run from a 9 volt source if the regulator is removed, and replaced from input to output pin by a jumper.

Download the manufacturer's data sheets for the MF100, LM555, and LM386 for the operating voltage ranges of these IC's, if you are interested.

73,
Bruce N1RX

Just got my NESCAF up and running. Seems to work great!

However, there is one "undocumented feature" that has me a bit confused.

I plugged the NESCAF into my Rockmite 40 and I was amazed at the fabulous performance of the NESCAF. Worked so well, I started calling CQ in hopes of a FB QSO. Sure enough, I contacted some fellow way off about 1500 miles to the east.

HOWEVER - about half way thru the QSO, audio went flat - a little bit of internally generated noise, but no signals and no static.

Fiddling with the two knobs did nothing, so I cycled power with the on-off switch and all was well again.

This has happend numerous times. Usually as I am in the middle of a QSO, - you know - pounding away with the paddles, and then BTU but no dice cuz the audio has gone away again. I then quickly cycle the power and try to guess what the other fellow's first word was. Once or twice the audio just diminished rapidly as I was "reading the mail".

What gives?

Anyone else ever experience this?

Carl
WA7CS
Pasco, WA

That's a new one on me, Carl. The situation you describe sounds as though something is failing after it's been running a while. Perhaps a poor solder connection somewhere.

If you have access to some test equipment like a voltmeter, etc you could do some tests to narrow the possibilities. For instance, when the audio seems to disappear, check the output of the 9 volt regulator. Is it still OK? Similarly, you could check to make sure power is getting to the appropriate IC pins when this happens.

Of course, an oscilloscope or frequency counter would allow you to see whether the clock generator (LM555) is still producing it's output pulses, and are they getting to the SCAF chip. An oscilloscope would allow you to follow the rig's audio signal through the filter's signal path.

I'd start be double checking ALL the connections, looking for cold solder joints, etc. Also check all the IC pins. If you used IC sockets, make sure a pin is not bent over, and just touching the socket not actually entering it.

GL,
Bruce N1RX

Thanks for the input - however I was hoping against hope that this was an oft-encountered scenario that could be cured with some simple and obvious magic.

Darn - I just hate it when the first suggestion is to check for cold solder joints. I guess that is probably because that is usually the problem.

Since this phenomenon is unheard of to date, I guess I'll just do what I knew what I'd end up doing in the first place - checking those solder joints.

Maybe I'll get lucky and find a bent IC pin. I'm glad I put those sockets in there!

Of course, I have already managed to shoe-horn the board into a teensy little tin. A perfect fit that will probably never be replicated.

For some reason, if I just clobber a project together it works like a charm the first time and forever more.

If discipline wins out over speed and a burning desire to get 'er done, I ALWAYS end up taking apart a beautiful bit of craftsmanship.

Taking it apart will give me a good excuse to add a nite-time bypass switch for the BCI resistor at the RCVR input. Right now I just poke the resistor into the ckt board holes when needed.

I wasn't kidding about the performance of the NEScaf - it really does work better than I expected.

Good stuff!

Carl
WA7CS
Pasco, WA

I have my SCAF going and it is very nice, I got an extra MF100 to make a notch filter but have stalled out. Has anyone built a notch filter (band reject) with this? I find some information on the web but I guess I need a paint by number sort of detailed plan.

Thanks for any help.

73 Jerry

Hi Jerry.

The datasheet for the LMF100 chip provides guidance on using the IC as a notch filter. I have not done it myself, so I cannot provide you with 'paint by number' directions at this time.

I will say that if you decide to set up a separate stage as a notch filter, in tandem with the standard bandpass filter, you should feed it with a different clock generator. Otherwise, the notch would be at the same frequency as your peak (!). It should have a separate clock so it's center frequency can be controlled independantly.

GL,
Bruce N1RX

I measured the resistance of the dual gang pot, and am reading 50.8K on the front section, but, only 46.2K on the rear section, as measured with a Fluke 110 DVM.
Will this cause any problems in the operation of the Scaf Filter? I don't think so, but want to be sure, before I mount it.
73, Bill, k6mgo

This won't be any problem with the performance of the Scaf, Bill.
-Bruce N1RX

I would like to use the NEScaf filter between the audio output of my Elecraft K1 and the input to my Soundworks PC audio amp/speaker system. From some preliminary experimenting, the Soundworks amp needs little drive and I think I might be better off if I eliminated the LM386 from the design. I'm not a designer, so I am wondering at what point in the schematic would it be best to "tap-into" the circuit and eliminate the LM386. I am thinking right after capacitor C9, but I'm just guessing.

I would greatly appreciate comments from the more advanced builders out there on my plan and especially the tap-in point.

Thanks and 73,

Hank/KEoCU

Hi Hank.

Yes, on the output of C9 would be the place to tap your output. You would still want the on-board volume control in-line, so you can set the proper level.

GL,
Bruce N1RX

Thanks, Bruce. Can I just not attach the components at all to the right of C9 (all the audio amp circuitry) and expect everything to work ok?

Hank/KEoCU

Yes, Hank. For your 'line level' ouput application, you can just leave off all the LM386 circuitry. Or.... you could put all those components in. Just put an 8-pin socket in place of the amp IC. That way, if you decide to use the amp later, you can just plug the chip into it's socket.

73 es GL,
Bruce N1RX

Thanks, again Bruce. I'm going to get mine ordered!

73,
Hank/KEoCU

I ordered my fifth NESCAF back on July 25. Wondered if you were back ordered, as the other ones came quickly. Building this one for a disabled friend.

John-W2AGN

Hi OM, I ordered the kit on 26 July and haven't heard anything about arrival of the kit. Is there a big delay in shipment? Thanks, Eric, W0OBF, St. Cloud, MN

Several club members planned to discuss the NEScaf at Boxboro last weekend. I haven't heard the latest, but expect some news soon.

73, Scott N1AIA

All orders have been fulfilled, as far as I know. If you are still waiting on a kit (or anything...) please contact me. Sorry for any inconvenience.
/dave nt1u

In putting my NEScaf kit together, I couldn't find any polarized caps labeled "105" per the wiring manual (March 2008). I did find 3 unused caps labeled with 1 microfared and the number "35". I assume these are OK, and I am going to use them, but I also wanted to check with the Pros. Comments? Thanks and 73!

Looking at the March 2008 version of the wiring manual, there are, what look like to me, some contradictions which has put my completion of the NEScaf filter on hold. They are all on the page which contain "Stage 3: Clock Generator" and "Stage 4 The Scaf Filter".

About R10, the last line of the first paragraph says "The right tab and center wiper tab on the pot are connected together with a short piece of wire." This is not shown on the schematic included in the manual. So which is it?

About R7, the 4th paragraph is very confusing as to which is the wiper terminal and where it is connected. We have "Take care that the wiper connection (middle pin on pot.) goes to the designated W connections on circuit board." But 2 sentences later "Looking at the front of the potentiometers, connect the right tab of each gang to the correct circuit board location labeled W." Huh??

I'm pretty sure I know what to do here, but this part of the manual doesn't make sense and I would like to have this cleared up. Any takers?

The three caps you have are correct. The "35" indicates the maximum working voltage for the cap. Just take care with polarity when you install them. There should be a small "+" symbol near the positive lead.
GL,
Bruce N1RX

Regarding the instructions for the pots- R10: It doesn't really matter if you short the wiper to the "unused" end or not. The pot in this case is being used as a two terminal variable resistor (or "rheostat"). In general practice, the unused end of a three terminal potentiometer is shorted to the wiper. This prevents the device from presenting an open circuit if the wiper should fail. In the NEScaf, it doesn't matter if you do this or not.

R7: This is the dual section "bandwidth" pot. You are right, this section of the instructions is, at best- confusing. At worst- incorrect.

The first two sentances are correct- The wipers, or middle tabs or each section go to the R7a_w and R7b_w connection points. (W is for wiper) The Third sentence is incorrect. The right tab of each pot section should go to the R7a and R7b connections, respectively. I have rewritten that paragraph below. Hopefully this is clearer.

"Panel mounted, dual gang 50k Ω potentiometer--Take care that the wiper connection (middle pin on pot) goes to the designated W connections on circuit board. The circuit board is labeled R7a and R7a_w for the first 'gang' or section, and R7b and R7b_w for the second. The W stands for wiper (middle tab on each potentiometer section).
Looking at the front of the potentiometers, connect the center tab of each gang to the correct circuit board location labeled W. The center tab (wiper) and the left tab are connected together with a short piece of wire. Next, connect the right tab of each gang to the correct circuit board location for each gang( R7a and R7b). So, if you connect the center tab (wiper) of the pot section closest to the shaft to R7a_w, then connect the right tab of that same pot section to R7a. Twist the two leads to each pot separately as it makes it easier to keep track on the PCB. Some components already mounted on the circuit board might ‘hide’ the W so be familiar with the PC board layout. "

Hopefully this will help sort this out, Hank.

73,
Bruce N1RX

Bruce, thank you very much for your post. I did end up wiring the components correctly, and the NEScaf is working, although I have yet to install it in an enclosure. Still looking for the right one.

73,

Hank/KEoCU

You know right at the end of assembly Stage 1, where it sez power it up, and then warns the builder to be extra careful regarding reversing polarity. Well, you guessed it!. I let the magic smoke out of the 78L09.
I am organising a replacement 78L09, but I am wondering if I have damaged any other components in the Stage 1 build. All the resistors appear unharmed, but I have no way of checking the capacitors. Anyone any thoughts ?

A very embarrased Wallace MM0AMV

I don't believe that anything could have been affected by reversing the polarity at this point. Until the ic's are inserted, pretty much everything there is immune. You may want to take a look at an article Joe Everhart wrote which is posted on the njqrp site

http://www.njqrp.org/quickies/quickie2.html

about various reverse polarity scenarios. In this particular application, a reverse diode in series with the power may be all you need...but I still recommend reading Joe's article.

cheers
/dave

Hi Dave,
and thanks very much for the info. Its put my mind at ease, and hopefuly when the replacement 78L09 turns up tomorrow I can press on with the build.
I am color blind and I suppose I really shouldn't build stuff, but I enjoy it and I am normally very very careful. I will put it down to a moments madness!
Thanks also for pointing me at Joe Everhart's page. I cannot track down IN6277 over here, but I have located the alternative 1.5KE18A. I will be incorporating that neat little bit of circuitry into future builds, and also doing a bit of retrofitting.

73 Wallace MM0AMV

For any UK readers, Farnell and Rapid have the 1.5KE18A at 1.71 pds stg each

Hi,

I am going to start soldering my kit but also need to get a box for it. I want something sturdy, doesn't have to be bulletproof like the Hendricks BLT cases but I would like something tough.

Any info on the right size box, name or model number, etc. would be appreciated.

Thanks and 73

Rem
K 6 B B Q
The BBQ Ham Shack
10 miles North of the Golden Gate Bridge

Hi Dave,

Well, the replacement voltage regulator finally turned up and the rest off the build went like a dream.
The filter works really well in fact much better than I had anticipated. This accessory is a keeper.
Thanks once again for your help.
I have put a couple of pics (without the burnt out 78L09) in the NEScaf Gallery.

73 de Wallace MM0AMV

John,
I have a k2 -10 watt use it for Qrp. How does the filter preform on SSB.? Have you heard of anyone using it for SSB
with the K2?

tnx
carl W3IY

Pages