FWIW - I used generic silicone pads on my EF3-3 Wolverine, drilled and tapped my own heatsink holes, I did de-burr them and can attest they are not 100% true and plumb. I have never had an issue as long as the transistor is tightened down securely and used a decent sized washer to spread out the clamping pressure. This is the 4th or 5th amp I've built drilling and tapping my own holes - the first ones were bad, but each time I get better at straightness (if that's a word)I did bend them at 90, though next time I will measure it.
The rest I did as you describe. I have experience with this step while building several FW amps. Not saying I'm deeply experienced.
Given my track record with grease, could I use Keratherm transistor insulators under the outputs and no grease? I know the guide says to use mica and grease, but just asking.
I will stress the most important thing is to de-burr the holes. Then tighten them down to the heatsink without the insulator, then put the PCB in place and solder them in, remove PCB with transistors and add the insulator and tighten down for the last time.
Not saying that mica-thermal paste isn't better for conduction, but I've been running my 3-pr Wolverine at +/-70V for over a year almost every day - and I push it hard at least once a week without an issue. Heatsinks stay warm, but never too hot to touch. Oh and it sounds fantastic!
Stuart,@Chiptech please confirm that you bent the legs of the output transistors at 90° then assembled the output transistors to the heatsink then placed the amplifier boards over the legs of the output transistors. Then finally soldered the outputs to the boards.
Also you need to ensure the large washers do not touch the edge of the pcb. As the full rail voltage passes close to the top edge of the pcb.
I've been thinking about your question. I did not "place the amp board over the legs of the output transistors." I will the next go around.
Thanks for that clarification.
Even if you used
You 100% must do this otherwise. If you solder them to the board first it's going to stress the board or the legs of your transistors and this would stop them from laying flat on the heatsinks. So that would account for your thermal paste not making contact correctly.I did not "place the amp board over the legs of the output transistors."
I use separated power supply for the input+VAS+predriver. Small toroid transformer+schottky rectifier+simple emitter follower based regulator. 10V higher than the power supply of the output stage.With any well designed class AB amplifier cap multipliers and CRC supplies just cause excessive power supply droop with no real advantage. The PSRR of the input stage does a much better job of removing any hum than either of those supply types would.
Did you watch Daniels video where he does the output install?Stuart,
I've been thinking about your question. I did not "place the amp board over the legs of the output transistors." I will the next go around.
Thanks for that clarification.
This is the only way to install the output transistors, outlined here and in the build guide. There is no other way to install/solder in the output transistors correctly.
Timestamped Here:
His comment about Q104 is because this was filmed using the old bias spreader implementation and now its Q103 as per latest build guide.
Last edited:
Whst do you advise me to use then? Just a C supply?With any well designed class AB amplifier cap multipliers and CRC supplies
Yes, a standard rectifier and capacitor supply works best. Some are using switch mode supplies but that can turn into a whole new learning curve quickly.
Hi everyone
I’m stuck. Any assistance from the experienced and knowledgeable would be amazing. I am pretty new to all of this but have learnt a great deal so far. I finished my 4th GB board IPS module and EF3 – 4(v2). I don’t have a bench power supply, and the dimming bulb doesn’t work with the Cobra. I powered it up with the Cobra S2, without outputs installed and with 0.5A fuses, and all was good. Outputs now installed and it is functioning perfectly.
The second pair IPS + EF3-4 boards, sadly a different story. R109 reading 500R between pins 1 and 2, inputs shorted, J103 installed, 0.5A fast-blow fuses – and powered up. 4/6 LEDs illuminated, D3 and D12 didn’t, and the fuses blew – see short video. On checking, Q8 and D12 were incorrectly installed. D12 was reinstalled with correct polarity. Q8 destroyed on removal (!) and replaced. Also replaced Q5/6, I think on vague suspicion alone. I was confident the problem lay with the IPS but, sadly not: when powered up I got the same result.
Next, I substituted the IPS module from the good pair. I got the same result exactly as in the video. When I returned this IPS board to the functioning EF3 – 4 it still works fine. So I figure the EF3-4 board must be where the problem lies.
Visually – all looks identical to the good board. Couldn’t find any deficiencies, solder bridges, obvious short circuits. Photos enclosed.
Diodes – all polarities correct. All polar capacitors correct. All transistors correctly oriented. All LEDs functional and orientated correctly.
I have DMM checked every component IN CIRCUIT. I appreciate the difficulties with this but anything unusual was compared with the functioning board wherever possible.
All resistors gave acceptable readings except unusual behaviour from 102, 103, 132A, 132B, but these were identical when compared with the working board. Capacitors, difficult in circuit but no discrepancies found.
I have checked every transistor B-C, B-E, C-B, C-E, E-B, E-C on both the functioning and the faulty boards. There are discrepancies in D109, and Q104, 105, 106, 107, 108. Q104 E-B is 0.986 compared with 0.696 on the healthy board; is this significant? All the other discrepancies are OLs on the faulty board v high readings on the good board. I suspect these are all because the outputs are installed on the good board, but I am not sure. All the values are on this spreadsheet Transistor values in circuit EP3-4.xlsx, discrepancies are marked in red.
If anyone can guide me as to where to go next, I’d be really grateful.
Jonathan
I’m stuck. Any assistance from the experienced and knowledgeable would be amazing. I am pretty new to all of this but have learnt a great deal so far. I finished my 4th GB board IPS module and EF3 – 4(v2). I don’t have a bench power supply, and the dimming bulb doesn’t work with the Cobra. I powered it up with the Cobra S2, without outputs installed and with 0.5A fuses, and all was good. Outputs now installed and it is functioning perfectly.
The second pair IPS + EF3-4 boards, sadly a different story. R109 reading 500R between pins 1 and 2, inputs shorted, J103 installed, 0.5A fast-blow fuses – and powered up. 4/6 LEDs illuminated, D3 and D12 didn’t, and the fuses blew – see short video. On checking, Q8 and D12 were incorrectly installed. D12 was reinstalled with correct polarity. Q8 destroyed on removal (!) and replaced. Also replaced Q5/6, I think on vague suspicion alone. I was confident the problem lay with the IPS but, sadly not: when powered up I got the same result.
Next, I substituted the IPS module from the good pair. I got the same result exactly as in the video. When I returned this IPS board to the functioning EF3 – 4 it still works fine. So I figure the EF3-4 board must be where the problem lies.
Visually – all looks identical to the good board. Couldn’t find any deficiencies, solder bridges, obvious short circuits. Photos enclosed.
Diodes – all polarities correct. All polar capacitors correct. All transistors correctly oriented. All LEDs functional and orientated correctly.
I have DMM checked every component IN CIRCUIT. I appreciate the difficulties with this but anything unusual was compared with the functioning board wherever possible.
All resistors gave acceptable readings except unusual behaviour from 102, 103, 132A, 132B, but these were identical when compared with the working board. Capacitors, difficult in circuit but no discrepancies found.
I have checked every transistor B-C, B-E, C-B, C-E, E-B, E-C on both the functioning and the faulty boards. There are discrepancies in D109, and Q104, 105, 106, 107, 108. Q104 E-B is 0.986 compared with 0.696 on the healthy board; is this significant? All the other discrepancies are OLs on the faulty board v high readings on the good board. I suspect these are all because the outputs are installed on the good board, but I am not sure. All the values are on this spreadsheet Transistor values in circuit EP3-4.xlsx, discrepancies are marked in red.
If anyone can guide me as to where to go next, I’d be really grateful.
Jonathan
Just some ideas;
D12 backwards shouldn't have caused shorting faults.
If Q8 was installed backwards, I think it would have created a short to G2. The last time we saw a short on a users rail here on forum - it created a voltage spike that took out the cap multiplier protection diodes. Check D111 and D113 and surrounding components. This will probably mean there is no rail voltage getting to your IPS even when you installed the good IPS.
Curious in your video D10 & 11 don't light either..
Oh I just looked at your spreadsheet - okay this goes deeper..
Rail reversal or Q107/108 swapped?
D12 backwards shouldn't have caused shorting faults.
If Q8 was installed backwards, I think it would have created a short to G2. The last time we saw a short on a users rail here on forum - it created a voltage spike that took out the cap multiplier protection diodes. Check D111 and D113 and surrounding components. This will probably mean there is no rail voltage getting to your IPS even when you installed the good IPS.
Curious in your video D10 & 11 don't light either..
Oh I just looked at your spreadsheet - okay this goes deeper..
Rail reversal or Q107/108 swapped?
Last edited:
@mainframe99 thanks for the prompt reply! I have just checked, Q107 is Sanken C43883A, NPN; Q108 A1859A, PNP. I think these are correct. I am certain I connected the Cobra correctly to V+ and V-; my cables are colour coded and I triple checked, plus the same result has happened on 3 occasions. Is this what you meant by rail reversal?
I would move to testing the IPS / OPS in isolation steps in the build guide. Are your predrivers metal backed like your cap multiplier semis? Sorry if I'm stating the obvious.
@mainframe No, the pre-drivers are fully encapsulated, not metal backed. I checked everything for shorts to ground, including between thr predrivers and the heatsink, but drew a blank. I don't mind anyone stating the obvious, it could be something obvious I'm missing! Re testing IPS/OPS in isolation - yes, I suspect I am just putting off the inevitable. Any recommendations for an adequate bench power supply?
A 42VAC transformer will produce about 59VDC. Normal mains supply allowance is +/- 10%. Normal rule of thumb safety margin is 10% so you should be using minimum 72V caps.

Please keep the commercial offerings in the Group buy thread.
- https://www.diyaudio.com/community/threads/the-wolverine-4th-group-buy.411286/post-7787059
Hi Guy's,
We're excited to share some great news that we hope will enhance your amplifier building experience.
We now have anodized heat sinks available for the Wolverine project.
Please check the Wolverine 4th group buy thread for details.
Happy building!
Best,
Stuart & the Wolverine Team
We're excited to share some great news that we hope will enhance your amplifier building experience.
We now have anodized heat sinks available for the Wolverine project.
Please check the Wolverine 4th group buy thread for details.
Happy building!
Best,
Stuart & the Wolverine Team
I am close to finishing my input boards iam in need of advice I have 4 fairchild ksc3503e that are all in the 170 range but only have ksa1381e in the 135 range i also have a bunch of tta and ttc004 both in 200 range the pnp is slightly higher but hfe spread is 10 or less for my best matches. Do i go with ksa ksc with 30plus spread or would the tta ttc perform better with higher hfe and closer spread iam wanting to use these for pre driver on main board and q7 q8 on input board i also have 10 original NEC 2sc2682 with hfe in 230 to 250 range is that too much for Q13 ?. I know stuart and fireanimal know could i get some of your expertise here Thank You in advance
Q13 Preference
1/ 2SC2682Y
2/ 2SC3503E
3/ KSC3503D
4/ TTC004
Q7, Q8
KSC3503 orTTC004
KSA1381 or TTA004BQ
No real preferred device, either is fine.
1/ 2SC2682Y
2/ 2SC3503E
3/ KSC3503D
4/ TTC004
Q7, Q8
KSC3503 orTTC004
KSA1381 or TTA004BQ
No real preferred device, either is fine.
- Home
- Amplifiers
- Solid State
- DIY Class A/B Amp The "Wolverine" build thread