SWTPC Tiger .01 207B Schematic

[PB2]

A member asked for this schematic, so here it is:
http://baselaudiolabs.googlepages.com/TIGER-01-207B.pdf

Hope it helps to repair the amp, I've never worked on one myself.

Pete B.

[PB2]

Here it is online, wish I had found this before I scanned it, LOL.
I think this is the A version and the one I scanned is the B, not sure of the differences:
http://www.swtpc.com/mholley/RadioEl...RE_Mar1973.htm

[Damon Hill]

Very good! I've got a couple of 207s here which I'll try updating. I also have a blank board which I'll stuff as a repair/experiment unit--possibly the last ever 207 to be assembled?

I installed MJ21193/4s in one and think I see a reduction in high-order distortion products. I can measure as low as .005% distortion at a one watt level, but of course it rises with power and tends exhibit significant high-order products above 5 or so watts. Careful adjustment of bias while monitoring the Tek AA501's output on a 'scope can help to minimize this problem; I'm curious to see what the change in the feedback network may have on this issue?

Anyone tried modifications to SWTPC amps, especially the 207 and Tigersaurus?

[PB2]

I scratch built several of the Tiger amps years ago as a kid and always wondered why many of them blew up. I've not looked carefully at the real hardware for probably more than 30 years now but I do plan to take a look, perhaps when I retire, LOL! Something that I always noticed was that the output signal would go very close to the rail, which means deep into saturation. This is because the driver has more voltage (Vce) to work with than say a typical EF type of configuration. This is good for using all the voltage available from the rail, good for efficiency lets say. However, BJTs become VERY slow when deep in saturation and I'm fairly certain that this leads to severe cross conduction and they might even become unstable due to the slow down. I've often seen the capacitors melt in the output stage. I've seen cross conduction peaks in simulation of 2 amps or so. These amps are dangerous for this reason and I would not use them unless you could be sure to protect your speakers in some way. Perhaps reduce the fuse sizes. I considered clamping the drivers/outputs to avoid saturation but was unable to come up with a simple, low parts count solution. A thought just came to me, perhaps reduce the output BE resistor from 100 ohms down to 22 or even 10 ohms, then use something like the MJE15032/33 as drivers that can provide the current to drive that low R. The low resistance might turn them off fast enough, but it would take some analysis/experimentation.
You're probably aware that most poweramps with BJT outputs are not intended to provide full power into clipping for extended periods of time at HF due to this very reason, certainly not above 20 KHz. This is because the cross conduction duration becomes a significant length of time as compared to the period of the waveform.
My experience has mainly been with the Universal Tiger and Tigersaurus both of which I built from scratch, I also built the kit version of the Universal and repaired it probably 3 or 4 times after it blew up.
I had a lot of hope for these amps over the years, but now after seeing the cross conduction I'm not so sure.

[Conrad Hoffman]

The only time I've blown up my home-brew Tiger is on the test bench. They won't stand big voltage swings at 20kHz or more. In music service the thing has been indestructible for about 30 years or so.

[PB2]

Yes our kit built version of the Universal Tiger also blew up on the bench but we were not driving it hard, just sitting there with a signal generator connected. It also blew up under music conditions when driven hard a bit into clipping. This same amp did not blow up again for 20 years or so when used in light duty probably never seeing more than 20W out. This was the stereo version with less heat sink and no thermal cutout. My scratch built versions never blew up, but both ran on slightly lower supply voltages so perhaps that is part of it also. I also had figured out by then that the 100 ohm resistors needed to be 1-2W. And I did not try to use them as bench amps, that is for sure.

It certainly is not a bench amp, LOL! But there is something seriously wrong with it given all the reports of blow ups.

People also talk of them blowing up when they plug or unplug the input forgetting to turn it off. I've seen one with a note on top to that effect, LOL!

Damon, did you notice this thread: Universal Tiger

Some of the SPICE work that I did can be downloaded here: Swtpc Universal Tiger Improved And Simulation
I would include Bob Cordell's spice model file and replace all of the bad models with his.

[PB2]

I just added a new LtSpice simulation file at that link that includes Bob Cordell's models. I the old simulation to include reasonable selections from Bob's file. The sim shows 3A cross conduction peaks when clipping at 20 KHz. The zip file includes everything that's needed to run the sim under a standard LtSpice install.

[sakis]

this is old in a way .... but here is what i ve seen

if you look at the original application notes or built info original as was for the tiger 01 it seems that the way is made especially regarding ground distribution , pcb , and wiring between transistors is a complete nightmare .

I am pretty sure that any amplifier constructed like this will have serious stability /oscillation problems .

It could be very interesting to look at the simulation results see if there is any trouble there and then clone the amp with modern devices proper compensation and a new pcb without all these problems .

since i don't run simulators PB2 here you jump in !!!!

Again i think that at the time nobody was cloning Tiger 0.1 or one may copy it as was so mistakes made where simply repeated so also here is how these amplifiers have a failure history ...

I rebuilt one of them as we speak but i am also trying to correct the building mess I will try to keep everything original except the building /ground / distribution errors

kind regards
sakis