Silicon Chip 200Watt LD amplifier

[mickyh]

HI Anybody else built or had experiance this amplifier with the ThermalTrak Transistors ??

I have just completed 2 of these 200Watt LD amplifier modules (Jaycar kits) i have one problem, even with the 47om resister in Q7s emitter i cannot get anywhere near 7-10mV across any of the .1ohm emitter resisters mine all read around 2.5mV. with 30mv across the speaker terminals, (Rail voltage is +55.8 , 0 , -55.8.
The voltage acros the diodes (Q10 & Q11 base) is OK at 2.28Volts. All other circuit
voltages read Fine as per the article , & the modules seem to work OK, (listening as im am writing this ) sounds pretty good !!!!
both modules are exactly the same , so what am i missing here??
Are there any updates to this article??
Component variation problem?? or what

Thanks Mick Hawtin

[lineup]

7-10 mV gives 70-100mA, which is normal

The diodes should create a higher voltage, the more current.

1. You can increase current in Q7, Q9. But not too much, cause they may overheat.
54V x VAS-current = power dissipation in those TO126, BF469/470.
With good cooling maybe we can use 2 Watt for those TO-126

2. A slight effect you could get from reducing BASE resistors of MJE15030/31 to 47 Ohm. (from 100Ohm)

3. Reduce current in drivers a bit will lower the voltage drop in drivers.
= increase EMITTER resistors of MJE15030/31, from 100 Ohm to 220 Ohm.

4. Parallel each 0.1 Ohm emitter, with another 0.1 Ohm, will double the current:
3.5 - 5 mV for ame 70-100 mA
-------------------------------------

These adjustments can all effect the operation of the amplifier.
How little or much is hard to say.


Here are some topics about use of Thermal-Trak.
With posts by clever amplifier designers like Bob Cordell & Douglas Self !!!!!!

Biasing/thermal compensation of Thermal Trak transistors
Biasing/thermal compensation of Thermal Trak transistors
On Semi ThermalTrak
On Semi ThermalTrak

Below is the schematic of this Ultra-LD 200w amplifier

[lineup]

Another idea, which would make other changes not necessary,
is to put in one resistor with a suitable value in series with those Thermal-Trak diodes.

I do think this could work well

[AndrewT]

Hi,
Lineup is on the right track.
4 series diodes cannot generate sufficient voltage drop to drive sufficient bias current through the output stage.
Adding a pot in series or another diode would allow a higher voltage and result in a higher quiescent current.

[mickyh]

Thanks for every ones feedback on this Amp problem , i actually contacted one of the designers , John Clarke who came to the same conclusion that you guys did,
that is ,the fixed resistor in series with the diode string to increase the quiescent
current in the output stage, i finally ended up with a 7.5ohm series resistor for
8.5mV across the emitters = 85ma .
Amp running fine now , very stable , hopfully this will help some other builders who
may have the same problem.

Regards Mick

[lineup]

Great!
Now we know it is not that difficult to adjust bias of ThermalTrak transistors.
Using the 'simple resistor trick'

[alfredrofe]

Originally posted by mickyh
HI Anybody else built or had experiance this amplifier with the ThermalTrak Transistors ??

I have just completed 2 of these 200Watt LD amplifier modules (Jaycar kits) i have one problem, even with the 47om resister in Q7s emitter i cannot get anywhere near 7-10mV across any of the .1ohm emitter resisters mine all read around 2.5mV. with 30mv across the speaker terminals, (Rail voltage is +55.8 , 0 , -55.8.
The voltage acros the diodes (Q10 & Q11 base) is OK at 2.28Volts. All other circuit
voltages read Fine as per the article , & the modules seem to work OK, (listening as im am writing this ) sounds pretty good !!!!
both modules are exactly the same , so what am i missing here??
Are there any updates to this article??
Component variation problem?? or what

Thanks Mick Hawtin

Hello Mick, my solution to the quest for emitter voltage was to place a trim pot in series with the diodes, on the board near Q 12, cut track and solder directly. After going through D Self's distortion notes on his website I also picked both ends of the 100R emitter resistors on Q10 and Q11, removed them from the SP output rail and joined them together, palcing a 1uf cap across these two resistors(from emitter to emitter) also decreases crossover distorion. on the second amp I removed both reistors and placed a 220R on the board close to q10&11 with a 1uf cap across it. this mod lifts a veil espceially in the mid and upper ferquencies. I was drawn to this fix and Self's site because someone mention it was a derivation of Self's blamess amp. He also suggested getting rid of the tail resistor in the LTP, I asked him why? he told me that it causes problems on power up when you use a variac. So I don't see any reason so far to remove it.
Hope this works for you.
By the way I blew up Q9 and was forced to replae it with a BF459, higher Ft but also higher Cob, the difference between channels is noticable, the BF469 with 1.8pf is better, this it mentioned in the book by Randy Slone.

[ostripper]

By alfred - I was drawn to this fix and Self's site because someone mention it was a derivation of Self's blamess amp. He also suggested getting rid of the tail resistor in the LTP, I asked him why? he told me that it causes problems on power up when you use a variac. So I don't see any reason so far to remove it.

Ran into small issues with my blameless design.. same as this
(almost..led CCS) get rid of the 6.8k 1W(jumper), Q5 Re=300R,
Q7 Re=100R... You will get less turn on noise, cooler running
VAS

By the way I blew up Q9

and a little more open loop gain. (1.8ma LTP/ 6.5ma VAS)this was measured on the real thing ,not sim.
BTW, a good stable design (ran it for a month before moving
on to new design)..
OS

[pheonix358]

The ULD Mk2 seems quite good. The second part of the blurb covered various things mentioned here. On a group buy these modules are around $AU50, $US33. Quite good value. Would any one like to offer advise on the following.

The amp uses Four thermalTraks while the manufacturers info shows six devices, probably why Leo and John ended up having to add resistance to the temp diodes. What happens if we put the other two back in.

Any thoughts on getting rid of the NP electrolytic input cap

Any other mods without changing the board layout.

Thanks

[alfredrofe]

Originally posted by pheonix358
The ULD Mk2 seems quite good. The second part of the blurb covered various things mentioned here. On a group buy these modules are around $AU50, $US33. Quite good value. Would any one like to offer advise on the following.

The amp uses Four thermalTraks while the manufacturers info shows six devices, probably why Leo and John ended up having to add resistance to the temp diodes. What happens if we put the other two back in.

Any thoughts on getting rid of the NP electrolytic input cap

Any other mods without changing the board layout.

Thanks

I would like ot know how to get them for $50?
I believe adding two other output devices would require some extraneous wires if you don't object to that and the possible side effects, it is perfectly amenable.
My habit is to replace all electro input caps with stacked polyester(1uf), probably because I have some in the cupboard, electros have a little hysterisis, sluggish in LF and poor mid and high freq quality/response, despite the fact that eclectros are getting beter over time.
Please refer to post 7+8, however this requires a little surgery on the board.
refer to other links
http://www.diyaudio.com/forums/showt...94#post1650894
http://www.diyaudio.com/forums/showt...23#post1651523
alfred