SKA GB150D now public domain...

Here are some specs.
Recommended Supply Voltage: ± 50V
Output power (± 50V regulated, < 1% THD): 150W at 4Ω or 8Ω
Voltage gain: 28
Input sensitivity: 1V RMS for 100W at 8Ω
Input impedance: 27K Ω || 330pf
Load impedance: 4 ohms or 8 ohms
Idle Dissipation: ~11W at ± 50V DC
THD 1kHz, 8 ohms load: 0.025% (100W), ~0.01XXX% Typical
Your voltage is a bit on the high side. Is there a switch to set different ac voltages on your amp?
 
Thanks for your reply bigun.
Yes I saw many opinions regarding setting bias, and I tried with 150 ma per device, couldn't really hear any difference, and heatsink got more hot than good, also when I start it up cold with this high bias setting it draws alot of current in 5 minutes before it stabilizes, I will aim for lower bias and longer lifespan.
I live in north maybe have a flip switch with summer/winter bias :).
Thanks again.

Well, I didn’t think the thermal control in Greg’s design was bad, but I certainly improved on it when I redesigned for my clone and with a far better pcb layout than his. (probably why he deleted my posts from his forum and locked the thread without any explanation!).
 
Here are some specs.
Recommended Supply Voltage: ± 50V
Output power (± 50V regulated, < 1% THD): 150W at 4Ω or 8Ω
Voltage gain: 28
Input sensitivity: 1V RMS for 100W at 8Ω
Input impedance: 27K Ω || 330pf
Load impedance: 4 ohms or 8 ohms
Idle Dissipation: ~11W at ± 50V DC
THD 1kHz, 8 ohms load: 0.025% (100W), ~0.01XXX% Typical
Your voltage is a bit on the high side. Is there a switch to set different ac voltages on your amp?

It does not have a switch to drop voltages for the ac. The transformer is just the Yamaha unit, 630va, and supplied t channels with 120wpc continuous. Looks like actual rail voltage from the yammie schemammie is +-58.3v. so yeah, kind of high.

So I am still quite a newb at this stuff, so maybe you can tell me if I am mistaken here. Higher rail voltages will lead to higher output by virtue of a larger signal. There will be additional heat generated by the higher voltage. If the heatsink/fan are able to control the heat, I should be ok, no? Would I want to make changes to the circuit or bom to account for the higher rail voltages?

Thanks for your help.
 
I included the protection diodes as they are a good idea and I had no issues with the final version, but I did include RC networks on the gate-drain as dictated by good engineering practice to forestall potential oscillations. I used a tight layout to keep the parts close to the MOSFET’s.


Sometime, I should try cranking the bias again, let ‘em run as hot as my heatsink allows and see how it sounds in case it becomes a contender for a high sensitivity speaker project where Class A would be nice to have, along with the dynamic headroom of switching up into Class B.
 

Attachments

  • 37CEAFB9-DB8D-4835-96F4-3A4C8D038122.jpeg
    37CEAFB9-DB8D-4835-96F4-3A4C8D038122.jpeg
    824.4 KB · Views: 586
So, moving forward, I have the boards, and have started stuffing resistors. I need to source a few parts though, and have a couple of questions.

Input transistors, specifically, the c2240bl and a970bl transistors. These are getting difficult to find, and need to be matched. BD ent has them for $3-4 a pop. The thread discussing alternatives for c2240/a970 indicates that the c1845 and a992 transistors are good alternatives. I have a bunch of these. Can these work in this design? If not, do you guys have a good source for finding these in matched pairs? Popular unnamed auction site? (ugh)...

Output mosfets, ideas to source well matched Vgs pairs of p fets and n fets? The p mosfets need to match each other, and the n mosfets need to match each other, it's not about matching complementary pairs. Is this correct?

Thinking about the power supply. I plan on using the transformer from the Yamaha. And given the layout constraints imposed by the heat sink location, I am thinking about creating two capacitor banks, 40,000uf each. Is this adequate for this design, or too high? If I am using two separate banks, I would be setting them up to operate basically as mono supplies (driven by common transformer). So two rectifiers, two capacitor banks, etc. I am learning that AB amplifiers don't necessarily need the CRC layout, and a straight capacitor bank can be used, and I have set my HB up this way. I believe the same would apply here as well, yes?

Thanks for the advice!
 
this maybe heresy, but the LTP’s in this amplifier are not at all balanced by design, matching transistors may not be critical. And mismatch produces a bit of 2nd harmonic, most of which will be compensated for by the symmetrical topology - in short, if finding transistors is a struggle, don’t sweat the matching.
 
My boards are the Jims boards, (I looked for the real deal from Greg Ball, but his site is gone...) So I am going by the schematic and BOM provided by Jims, as posted earlier.
I believe I am after these mosfets IRFP240/9240. The schematic shows the BC parts, but it is called out to use the c2240/a970 transistors.

Thanks guys!
 

Attachments

  • GB150D BC part list v1.1 (2).pdf
    58.9 KB · Views: 159
  • GB150D Sch v1.1 (2).pdf
    65.7 KB · Views: 162
I have not made much progress on this, but I did finish up soldering the boards, except for stuff that needs to mount on the heat sinks. Thanks Jeff, I went with the a992/c1845

Where should I buy matched mosfets irfp9240/irfp240 for the outputs? Evil auction site? Diyaudio mosfets for the f4? Anybody holding? Ha.

Also, with the 58v rails, what sort of capacitance should I look for in the filter caps? I was planning on using 63V caps thinking 4x10,000uf, but am worried that may not be enough. 4x10,000uf for each channel? If I can get away with 20,000uf per channel I will pilfer caps from an arcam avr600. I think I will build each channel with it's own rectifier and capacitor bank. I've been looking for a page that will let me calculated needed capacitance, but struggling finding one.

Zobel networks, this amp does not have it on the board, is it needed with this design? I can integrate at the speaker outputs if needed.

Thanks guys.

Proof I am making a little progress ...
 

Attachments

  • IMG_20200923_224707.jpg
    IMG_20200923_224707.jpg
    982.2 KB · Views: 625
Last edited:
Made a little more progress. I'm going with 40,000uf in a single bank of capacitors, decided to use the power supply from the arcam I previously turned into a honey badger. I set it up with 6 terminals at the star ground.
I think I am going to order the mosfets from the diyaudio store. I think I will be able to integrate the speaker protection from the Yamaha, but unsure. More research. There will be spare +-45v rails but I think they are low amperage. Researching their use in the Yamaha, see what I can come up with for a use. Or just leave secondaries disconnected.
The Denon badger sounds great, and was easy, but was not as satisfying. Too easy. Getting excited about this one. I love integrating different pieces together like a puzzle. We'll see how it ends up I suppose.
 

Attachments

  • IMG_20200925_235516.jpg
    IMG_20200925_235516.jpg
    117.8 KB · Views: 545
  • IMG_20200925_235533.jpg
    IMG_20200925_235533.jpg
    104 KB · Views: 475
Not much activity in this thread. I am still making progress after the kids go to bed. Power supply is 100% done and installed. I made a custom plate out of aluminum to mount it. Mounted up the amp pcbs on the heat sink. I ordered the fr mosfets from the diyaudio store, so when those arrive I will finish up the amp part. Last piece will be a dc protection board.
 

Attachments

  • IMG_20201002_183128.jpg
    IMG_20201002_183128.jpg
    372.5 KB · Views: 128
  • IMG_20201002_180026.jpg
    IMG_20201002_180026.jpg
    113.6 KB · Views: 111
  • IMG_20201002_180030.jpg
    IMG_20201002_180030.jpg
    95.5 KB · Views: 387
  • IMG_20201002_180354.jpg
    IMG_20201002_180354.jpg
    148.3 KB · Views: 394
Alright, could use a hand here. I've hooked it up. I put 10r 1w resistors in place of the fuses to the output. I shorted the inputs. Outputs are not hooked up. Fired it up with a 75w dim bulb tester. No smoke came out, so that's good. But it is drawing too much current, and the rails aren't rising past 20v. Dim bulb tester is lighting up, but not full light, maybe half brightness. Rails when it's in the wall are +-58v (did not power up the boards without the dbt).

Vr1 is for adjusting dc offset, I measured dc offset of -120mv on one channel, and -60mv for the other. Not a terrible starting point there. But measurement is stubborn, and adjusting does not have a lot of impact, and I don't want to overdo it. Which way should raise the dc offset? Counterclockwise?

Vr2 is for adjusting bias. But I am unsure where to measure bias at. I was grabbing it from the drain center pin of the outputs on either side. One side was showing 0.0v, and other side read 7v. So if I am calculating bias current, it would be 7v/20r = 350mA. Am I measuring and calculating my bias correctly?

I've searched for a guide in setting it up, but been unable to find clear instructions. My variable resistors don't click when they reach the end, so unable to start from center. What is the right way to set this up? My boards are the Jim's boards built to bom spec except for inputs. I'm using a992 and c1845 per earlier discussions. They were not matched, but from adjacent cut tape. Where is the right spot to tap for measuring bias? 10r 1w across the fuse holders is correct yes? What voltage should I be looking for across those resistors?

I've attached a PDF of the schematic I am working off of.

Thanks!
 

Attachments

  • IMG_20201007_234250.jpg
    IMG_20201007_234250.jpg
    147.4 KB · Views: 163
  • GB150D Schematic v1.1.pdf
    65.7 KB · Views: 164
Last edited:
So, rather strange, the side that was drawing the current and had 7v across the 10r resistors in place of the fuses is now all set. It took a bunch of turns on the pots, but I got the offset to sit at 6mv. I was able to adjust to 320mv across the r10, so bias would be drawing 15mv across each 10r. I put a signal through and tested with the scope. Nice and clean. So I popped fuses in (still all hooked up to the dim bulb tester) and it sounds very nice with my just test speaker. That side is ready to go off the dbt and get adjusted on the wall plug.

The other side, I can't seem to get any voltage across 10r at the fuse holders. 0.0mv. nothing budges it. I did get the dc offset to set at 9mv. But it required tinkering with both pots to find the spot. But I need to figure out why I get 0v across the 10r. Any ideas?

Thanks. I am excited to see this one coming together, this has been a fun build so far.