February 1, 2020 at 5:51 pm #10493
I started building this pedal awhile ago. I would have posted it under the “Show Off Your Build” heading, but this does not use a Guitar PCB board. The pedal was built to mimic the built in fuzz circuit in Mark Farner’s Messenger guitar that was used on the first few Grand Funk RR albums. The Heathkit TA-28 fuzz was rumored, although likely incorrectly, to be the fuzz circuit in Farner’s guitar. Versions of this fuzz seen on the internet sound very close to the Messenger guitar fuzz. I used this fuzz’s known circuit as the template for building my fuzz, although mine is of course modified as to be a somewhat different circuit.
This build was also my first attempt at a veroboard build as I had never used vero to build anything before. It has been a learning process because I could not find an example of my exact circuit or close to it on the Tagboard site.
I learned to figure out how to transfer the schematic to Vero, although screwed it up a little, and learned from my mistake. I tried to design it for the least number of spaces and cuts, and I think it came out okay for the first try (I previously had done some things with perfboard) so I have had some experience transferring from schematic to a board of some type. The reason I wanted to do vero was so I could use sockets for the transistors.
Of course I had the circuit on a breadboard first to make sure it would work as advertised. However, it only sounded like a mild clean boost at best. I finally switched out the Q1 transistor which was supposed to be a 2N3391 with a 2N4401 and the fuzz commenced. I read somewhere the old 2N3391’s had lower HFE’s and mine was 360-370. I substituted the lower HFE 2N4401 with an HFE slightly below 300 and the circuit sounded like it was supposed to. Others who have tried the original circuit used a Q1 with an HFE of ~ 300.
Once I got everything off the breadboard and transferred to the veroboard, and then placed it into the enclosure connecting all the switches and pots, the circuit did not work at all, completely dead except for true bypass. Not to be daunted, I commenced to use my somewhat limited debugging skills to try and diagnose the problem. At first using a continuity tester, I found that the connection between the base of Q2 and the collector of Q1 was sporadic, which upon inspection was due to a bad solder joint that was reflowed and then allowed a perfect connect between Q1 and Q2.
Being paranoid, I then used the continuity tester to check all connections on the board, pots and switches twice, finding no other problem. So back into the box it went, and nothing really changed except that it now sounded like a weak clean booster with the circuit activated via the footswitch. So, I started checking the voltages against those in the original Heathkit schematic and found that they were way off. Since the circuit continuity was already tested and confirmed, I double checked all components for correct values and everything (resistors, pots and caps) were dead on. So I then checked the circuit for correct placement of components on the board with my diagram of the veroboard and everything was correct. Knowing something had to be amiss, I checked the veroboard diagram against the circuit diagram, and voila I had made a stupid mistake placing a resistor after the collector of Q2 instead of before the collector of Q2. Fortunately I had screwed up the spacing between some components leaving an extra space between Q2 and the 10K pot, so I placed another transistor socket in this location and moved Q2 over. I left the old socket in place because it would make no difference in the circuit if I took it out or left it in. I did not want to risk messing something else up by trying to remove the old socket, which was in a tight space.
I was then able to desolder one leg of the errantly placed resistor and angle it so that it now was placed before the collector of Q2. Now all the voltages measured to be within the proper range. The pedal sounded like it was supposed to!
The only problem was that the Volume and Tone pots worked opposite one would like – turning clockwise lowered the volume and decreased the treble. However the pots were wired as in the Heathkit schematic. Having some experienced with this type of problem, I just flip flopped two wire positions on the pots and now the pots act how I wanted them to act.
By the way, this is a 1.5V circuit run by a single AA battery, so I was unable to hook up an LED. So to make sure the circuit is off, I put in a Kill Switch to disconnect the battery power when the guitar cable is plugged into the input as below. Note that the Volume works more like a Gain or Fuzz control and is straight from the original schematic. The output control (10K pot) was placed in lieu of a 10K resistor that actually acts more as a Volume control.
Gutshot. You can see the empty transistor socket left in place.
Incorrect veroboard layout. The 8.2K resistor should be placed in front of the collector of Q2 (a PNP 2N2907 transistor).February 2, 2020 at 3:17 pm #10561
Correct Vero layout. I also should note that this is a somewhat noisy pedal, the same as others have observed. I would have liked to put some sort of filtering in the circuit to try and abate some of the noise, but that is way beyond my expertise.February 2, 2020 at 11:54 pm #10581BarryKeymaster
I would love to hear this.
Sorry I missed your earlier post. I saw it and there are a few bits of discussion involved. I apparently have the Flu so I am not doing so good at the moment. I will let Wilkie know about this.
Thanks for sharing.February 3, 2020 at 5:14 pm #10593wilkie1Moderator
This circuit appears to be using a NPN and a PNP transistor in a complementary feedback pair configuration known as a Sziklai pair. This produces a high gain circuit.
I would be interested in hearing an audio sample of your resulting pedal.
I notice that you have 2 10K resistors instead of 2 1K resistors attached to the tone pot. Also you have a 10K attached to the collector of Q2 instead of a fixed 10K resistor. Otherwise, the circuit seems to be the same as the TA-28.February 3, 2020 at 8:59 pm #10594
Robert Paul – I’m curious why you believe moving the 8K2 resistor has any bearing. As far as the electric\electroninc potential goes, I discern no difference between the first stripboard image and the second – and fail to nee any electric potential different between where it is on the layout – electronically, its the same schematic-wise. It’s still connects at the junction of the collector of the 2nd transistor and lug #2 of the “LOUDNESS’ pot. Does have more to do with layout than the actual components for the layout to avoid proximity interference?
And yes, the old Heathkit TA-28 fuzz circuit is noisy as the circuit is a high-gain construct.February 14, 2020 at 6:11 pm #10786
Sorry I haven’t responded back right away. I have had a bad upper respiratory illness that finally has started to improve with antibiotics, and maybe this was the lite version of the flu as I had a flu shot, but this year it has only been about 50% effective (attenuated the illness without eliminating it altogether). And then we had new floors installed with 5 days of moving furniture, etc. out of the rooms, 3.5 days for installation and then 6 days putting everything back and rearranging things somewhat. So this is the first chance I had to respond.
Wilkie wondered about the two 10K resistors instead of two 1K resistors attached to the tone pot (and yes it appeared to me that this was a Sziklai pair circuit).
From my research on this circuit, others who have tried it had problems with the output signal getting lowered a little when the effect is switched on, and that the signal was below unity. The proposed solution was to raise the two 1K’s to 10K, and a poster did that which brought it up to a usable volume, keeping the tone the same.
Wilkie stated: “Also you have a 10K attached to the collector of Q2 instead of a fixed 10K resistor.”
Another suggestion to increase output was change R8 (10K Ohm) lower, such as 3K or even lower, although this will start to change the intrinsic asymmetry, apparently. Someone noticed that as the resistance was lowered, there was a slow increase in volume along with some extra highs. The last few percent down to 0 (Zero) KOhm gave a sharper treble & volume increase. Some of the blattyness disappeared at 0 KOhm. So it basically gets nastier as the resistance is lowered but the overall tone doesn’t change too much at all. The thought was to highly consider a 10K pot there, especially since there is no volume control. Hence my “Output” control. The controls some of the volume and fuzz characteristics.
To address Cybercrow who was curious as to why I believed moving the 8K2 resistor had any bearing. To me it appears different schematically, as below (mistake on the left and correction on the right). Maybe I am wrong, and as Cybercrow suggested that electronically, its the same schematic-wise. It’s still connects at the junction of the collector of the 2nd transistor and lug #2 of the “LOUDNESS’ pot. It would be informative to me why there should be no difference between the resistor placement.February 14, 2020 at 10:04 pm #10792
Robert Paul – in the schematic you most recently posted, the left schematic does not match with either of the vero layouts you previously posted. The schematic on the right is what both vero layouts represent.
The left schematic shows a true “Sziklai pair” while the schematic on the right is not a true “Sziklai pair”. A true “Sziklai pair” has no other components between the two transistors.February 15, 2020 at 10:44 am #10799
Cybercrow, you are right, I made a mistake in my haste to redraw the corrected vero board layout. I know what I did on the board as I had it in my mind what had to be done to correct the circuit, but for some reason I didn’t translate it into the drawing properly. Below is the corrected layout. I believe this follows the schematic correctly.
With regard to a “Sziklai pair” I screwed up my sentence in my reply to Wilkie. I forgot to put in “like” a Sziklai pair “but not quite.” That is what happens when your wife is talking to you when you are trying to type, and I lost my train of thought. A Sziklai pair is similar to a Darlington pair, but instead of two NPN’s, it utilizes an NPN paired with a PNP. A Sziklai pair can be used to make a negative ground Rangemaster with a low gain NPN and PNP pair. If this was a true Sziklai pair situation, the transistor portion of the schematic would have to appear as below.
I agree on both of your comments. And I know you know what both Sziklai and Darlington pairs are. I put my comments in for those who may not know what these are (although this may be basic knowledge for most reading this forum).February 17, 2020 at 9:35 am #10890
Robert Paul – Fair enough. I fully appreciate the distracted typing bit. And to be honest, electronics is full of tiny details that can easily cascade into drastic end results. In any case, thanks for sharing your experience with this build. I love the old circuits that used such low voltage supplies.February 17, 2020 at 12:12 pm #10898
Thanks Cybercrow, your comments are greatly appreciated, which I look upon as POSITIVE criticism. I enjoy discussing the old circuits and that is pretty what I like building from scratch, but with some modifications. I wish others would jump into these discussions as I quite enjoy them and they are a learning experience for me as I am not by any means an expert in these type of circuits. I only have some limited knowlege of circuits from my college and post grad physics courses, but I enjoy using what little knowledge I have in trying to figure out what is going on in the circuit and attempting some simple mods.
Now on to finishing the modified bazz fuss that I have bread boarded and getting it onto a GBOF board (along with some additional vero and/or perfboard circuitry) and then boxing it up.
When I get some time I will attempt to make an audio recording of the pedal, but that may be awhile off.April 4, 2020 at 2:33 pm #12212
I am revisiting this topic due to recent discovery. While I was trying out different transistors in a modded bazz fuss I built, I discovered why the 2N3391A transistor didn’t work in the circuit. It is because it has a non-standard pinout for a 2N**** TO-92 package transistor. I didn’t look carefully when testing the transistor with my Atlas DCA 55 checking the HFE at the time of my original build. All transistors worked in the bazz fuss except the 2N3391A, so I retested it with my Atlas DCA 55 and being suspicious, checked the pinout. It did not have the typical EBC of a 2N TO-92, but instead it had an ECB pinout, confirmed by checking with available datasheets on the internet
I had checked out a 3391 pinout previously on Small Bear’s site where they had a “house” version of the transistor and it had the standard EBC pinout typical of a 2N TO-92, and that is why I got fooled. I could have modified my vero layout had I known this ahead of time, but I can’t find an easy way to make it work now. The 2N4401 at Q1 works fine and has a similar gain bucket as an old 3391 anyway. Learned another lesson – always double check the pinouts of transistors previous to designing a vero or other circuit layout.April 5, 2020 at 1:59 pm #12227
Regarding my post above, the pinouts are as below. The SB transistor is on the right.
- You must be logged in to reply to this topic.