Hartke HS1200 - what keeps the bias V in the middle ?

[chazpope]

well I did measure the current at TP2 (forgot what it was) but not knowing what value to go for I proceeded to setting it 'by feel' - so that with no heatsink the outputs do not get hot, and the lamp does not shine - I guess it likely draws 5W at idle.

I left the amp at the pub so it won't be until Sunday that I can measure and report it for science sake

Like I said I idled it for some time and it did not get hot, but not sure if it does not go into some other state after some real play. I got this amp in a cooked state to start with - I replaced 1 shorted output pair and that fixed it but there was no apparent reason why it had died. Since that I have played it for about 200 hours at lowish volumes until it burnt again. The bias has always been set to the lowest setting.

So it seems to me that the heatsink does not work when in kickback state - there is absolutely no air movement in that hole and even 20W of power makes it really hot, especially in a hot pub. The only proper test for this is by generator which I don't have right now.

[chazpope]

How about measurements.

So I took the amp back from the pub and I have measurements:

Voltage at TP1 and TP2 is 0.5 mV, idle consumption is 25W

Turning the bias pot full up results in TP to 13mv and power to 40W.

I measured AC current and I am not sure how valid this is to measure power. I can't see how 20W is dissipated so it might be less in fact.

There is no heat build up at idle.

[chazpope]

so have you any comments guys? The drivers have arrived but I am wondering now whether I should break it up again, remove the heat sink etc just to replace them.

[Jazz P Bass]

I forget what you are trying to fix.

[chazpope]

I forget what you are trying to fix.

well you asked for measurements - I gave measurements. Question was about how to set the bias, and the amp works only I do not have a reason why it blew up after only light use and why it gets so hot when played

[Jazz P Bass]

O/k.
That 5-13 mv reading sounds really low.
Like it's pointing to something is wrong yet.
As it is, that will result in very low mains current draw & low idle temperature.(which is good in a way)
Do you have a digital thermometer.
It would be nice to know what temp you are calling 'really hot'.
I would try to get a reading directly across each & every ballast resistor.
That will tell what each transistor is doing.
They should be relatively the same.
Is the TH1-TH2 (R3) connection good. (do you read 500 ohms)
Is this part firmly afixed to the heatsink.
This part is supposed to help track the heatsink temperature.
Voltage readings at the base of the output & driver transistors would help. (static Vdc, no signal)

[52 Bill]

Reading some of your earlier posts, sound like you were setting the bias with the amp still plugged into the light bulb limiter. If that is true, remove the limiter and set the idle bias again.

And also be sure to allow the amp to run until it has warmed up and become stable before biasing.

[chazpope]

Why do you think i13mV is too low ? It produces 10W of idle outputs power.

Nobody has yet said what the bias should be. Again, I get no heat at idle, no audible distortion at low volumes. No I was not doing bias with the lamp in.

I have a thermometer but how do I get it hot - it ain't easy playing two hours bass to the neighbors, I don't have a suitable dummy load nor a generator. Do you mean that I should work up the amp to a temperature and then set bias ? I guess I can do that when I take off the heatsink , might be dodgy.

I am opening it again to replace drivers, hope it's not for science only...

An hour later I have new drivers in and measuring voltages (with lamp)- rails 50V, output bases .55V driver bases 1.13v, wildly variant emitter currents : Q217 ans Q214 want to run away while the others stay cool. I guess this is normal with no heatsink still did not expect such a wide variation. so emitters are between 3mv and 15-20mv depending on temperature.

now I am not sure what happens with heat sink on - it will be harder to measure/touch.
also the pcb has just about had it but does it make sense to swap and put the HOT transistors next to the Thermal probe - I really don't want to pull them again

[Enzo]

It should run forever at the coldest bias. The only negative is that the crossover distortion would be the most. If you are not hearing any crossover distortion, then that is moot. If the amp sounds good, I see zero reason to set it hotter.

DO NOT POWER UP THE AMP WITH THE HEATSINK REMOVED. JUST DO NOT DO IT.

[Jazz P Bass]

If I am reading this correctly, the amp functions fine.
It simply gets really hot.
I asked for a heatsink reading when it's too hot.
The outside of the heatsink is accessable to probe.
It may be that what the OP feels is too hot, is not.
One of the original questions was' if the amp was only used at moderate volume, why did it initially fail?'

[chazpope]

ok it will be hard to measure that temperature, let's say it was too hot based on my experience of how hot amps get during play. This is not exact of course but let's say hot so that you cannot hold it indefinitely, I expect this was some limit T that it should go to after being on full (120W?) power for a while.

I wonder if anyone ran this at the rated 120W and for how long.

Enzo I see your point now - I had not seen this before (runaway at idle) but I have quick T sensing fingers so I spotted it in time Changed the drivers too almost killed - as I put the heat sink back on one of the tired track on a driver base had torn off and it looked dead for a while until I took it off again and fixed - it is still alive thank god

I will test this tomorrow and at the sunday gig, might try the temp gauge or my hands again and one day if I have a gen I will test what effect the kickback position has on heat dissipation.

[Enzo]

Sure, but the why did it fail question is usually one that never finds an answer. That question usually resolves into the asker assuming there is a weakness in the design, or that some particular action or use caused the failure. Both those things are usually not the case.

Many failures are due to random component failure. In the world of physics, there may be some direct trigger for the failure of a part that had a minor flaw. In other words a part is already compromised at birth, and some peak just happened to push it over the edge. Similarly, parts can be damaged or stressed by something and ultimately cause the part to fail days, weeks, months later. Like an oversize truck crossing a small bridge can crack its frame, then 3 months later it collapses when a Honda Civic crosses it. Not the Honda's fault. A solid state amp is one big feedback loop, so it is very hard many times to determine what part started a cascade failure. But even when we identify that first part to die, then we still don;t know why. For all was know, the pick and place machine put a tiny crack in a resistor when it stuffed the board. Over time with enough thermal cycles and vibration eventually the part gave way and burnt up. No one would ever know.


The amp may well still have a problem, but if the amp sounds OK with the bias all the way cold, I still see no reason to run it hotter. It SHOULD be happy as a clam to run that way. If it isn;t we need to find why not. Biasing it hotter would only mask other causes of overheating. Just my opinion, I am open to contrasting views.

[Enzo]

Nobody runs an amp at its rated power for a length of time. Not unless their music is a steady sine wave at max volume. A sine wave is great for testing, but a totally unrealistic signal for the amp. Play at that power level is a bunch of pulses with space between them. Watch a VU meter. It is only making 120w when it is making the loudest sound it can make.

remember too that your volume controls are not power limiters, they are merely sensitivity controls. If you play with the volume on 6, then turn the amp down to 3, you can play the guitar louder and come out the speaker just as loud as before. So the knob setting is not a good indication of how many watts the amp is producing.

[Jazz P Bass]

As an aside, a number of enclosed amps rely on the speaker movement to push air around in the enclosure.
Maybe at a lower output power setting you have found the 'doesn't do enough' spot.

[chazpope]

No of course the knob does not show the power, I do it all by ear - to start with the speaker (was it 70W) is totally inadequate for the power of this amp an it starts to distort fairly early. Say fine you plug a proper cab if you are going to play loud. Now say the amp plays audibly well up to power level X , I mean I use it at audibly half that level. It sounds to me like a 15W bass amp on full.

Re the power rating when an amp says 120W`I take it this means it can run a sinusoidal power of 120 reasonably indefinitely - of course not every manuf follows this I guess. So there is got to be a comparative measure and isn't that quoted what rms power stands for.

So tonight we did the blues jam with this amp in the same pub conditions. In the kick back state, 2 hours of play at moderate volume, ambient T 25C - and it got hot so that I could put my hand on the heat sing and hold it for no more than 5 secs. I say this is the hottest it should ever get on full power, I think I will try it just upright next time and see if there is a difference.