Hartke LH 500 blowing fuses.

[catstrat]

I've confirmed it's in the output section and i have no schematic yet. I have a variac so i can raise the voltage to about 20 volts at the variac to do some voltage test but it gets hot. What voltages should i have at the base, emitters and collectors of the o/p transisters? I know it's not going to be full voltage but will i still have positive collector voltage at 1 trans bank and negative at the other. Thanks.

[mozwell]

If your variac is getting hot, you are drawing way too much current. Something is shorted.
unplug the power amp (connector CN501-A)
First, check all power supply diodes, none of these can be shorted.
If all is ok, slowly run up on the variac, if possible, measure mains current, you should have less than 0.5A
measure all the power supplies, are they at the correct voltages.

power off, measure all power amp transistors, especially the output & driver transistors, all ok ?

[catstrat]

Hi Mozwell. Here's what i have so far. The first post i did was a quick post. I already tested it with the molex connector unplugged from the output module and it's fine so it's not the rectifier or transformer. I plugged it back in. The variac amp meter shows about 4 amps, too much, at only about 20 volts ac. Something is definitely shorted. The variac itself is cool, it's the heatsink of the O/P section that gets hot. I was able to get some voltage test done. This has 4 o/p transistors on each bank for a total of 8. The collector voltage of 4 of the transistors shows a negative voltage and the other 4 show a positive voltage. Keeping in mind it's not at full voltage,i remember it as about 10 volts respectively at the collectors. The emitters were at a negative .10 volts except one which was at positive .10 volts. I removed that one from the circuit and it tested as shorted. I left it out and tested the amp but it is still pulling too much amperage. I didn't write the exact voltages down, i'll check them again tomorrow and have more definite values. I checked the voltages again and this time i was getting a negative voltage at all the o/p transistor pins. WTH. I'm measuring the voltage with my meter grounded to the chassis, is this the right way? There is a small p c board about 1 inch square that has some surface mount stuff on it thats mounted to the main o/p board. What's that board for? Is there an easier way to test the output transistors without removing them from the circuit? I do a lot of tube repairs but transistor stuff is a mystery to me. I welcome all help. Thanks so much.

[mozwell]

ok, each of the big 2SC5200 & 2SA1943 transistors has three legs, with the legs facing "down" from left to right, its base (B)- collector (C)- emitter (E)
Power off, get a multimeter, measure on ohms between B-E, B-C, C-E, also do the same on the diode range, using both polarities of meter leads, ie red to B, black to E & black to B, red to E etc.
Ohms shoudl read non zero, diode range should read non zero. Check all 8 power transistors
Also do the same for the driver transistors, Q309 & Q310, these have the same pinout.
We shoudl also check the Vbe mulitplier (bias) circuit around Q308 is ok, Check that R319, SVR302, R320, TH302, D303 & D304 are ok.
We can put a temporary short across C-E of Q308. If there is a problem with this circuit, both output transistors will be driven full on, so to "disable" this circuit, we can just short C-E of the transistor as a temporary measure. Also check there is resistance from Q308 B-E, if the trimpot lost its wiper conenction, if may produce this fault.
Also check all the electrolytic caps look ok, are any bulging ?

Is the unit in bridge mode, if so, set it to stereo mode.
Unplug the wiring from the power amp to the speaker outputs, there may be a short there

My guess is, that we have a big short on the pwoer supply somewhere, that is drawing the big primary current.

edit.... also to check, from the middle pin of each transistor, is there zero ohms to the heatsink.
If all these transistors are mounted on one heatsink, they need to be insulated, otherwise we are shorting all the Collectors together, effectively shorting the + & - supply rails.

If any power transistors are blown, replace the whole set, positive & negative, as even it is "tests" ok, it will have been stressed and may not be reliable.
If you cant get the 2SC & 2SA transistors, MJL21193 should be able to replace 2SA1943 & MJL21194 should be able to replace 2SC5200
If any power transistors are blown, it is good practice to replace the drivers as well.

edit again
when testing the power transistors, also check their emitter resistors, on ohms multimeter range.
R324 to R331, all 0.22 ohms 5W. These all need to be ok, or the power transistors will not share current, and teh amp will eventually blow again.

If R23 is ok (10 ohm) then measuring to any volts to chassis will be ok. You can also use AN104A pin 2 or 3 as 0V for the power amp, or the end of R23 that connects to AN104A pin 2 & 3

[catstrat]

Thanks Mozwell. So you can check the transistors with a meter while they're in circuit. I'll check all these things today and get back with you on it. I have an amp repair shop here in Florida and i stay pretty busy but about 96 per cent of my work are tube amps. I turn down alot of solid state work because i don't have a good understanding of the circuits even though i understand what each componant does individually, it's when they are in a circuit and they're working with all the other parts that confuse me a bit. I can handle problems in solid state pre-amps with no problem but it's always the output section problems that just kick my butt. I know that we have the o/p transistors, drivers, pre-drivers and then the big mystery to me is the bias circuit. I'm guessing it's a type of a voltage divider bias circuit. Don't most of these have like a bias monitoring circuit to trip a fault light if the bias is out of wack? When i look at a tube amp schematic i know what each part in the circuit is for and what it does. When i look at an transistor output section schematic i'm just kind of shooting in the dark. Again, thanks.

[mozwell]

Hi
Dont be scared of solid state, there are many kind people here who can help you through any problem you may have.

yes you can get some basic information when checking power transistors in circuit. Basically we are checking for a shorted device.
even if only one is shorted, i woudl replace the whole set, positive & negative, as the shorted device will have "stressed" the others & they wont be reliable.

Very few (if any) have a circuit to monitor this "bias" circuit, and none have any LED indication. There may be a "clipping" LED, but thats about it. Generally even if all the output transistors are blown, the bias transistor will still be ok.

Bias for solid state, lets see if i can put this simply.
First, the power output transistors need (say) 0.6V from B to E to turn them on. For the top transistors the B is 0.6V more positive than the output, and for the bottom transistors the B is 0.6V more negative than the output. All htese power transistors provide the power to the speaker & the speaker current flows through them. These all have a current gain, so B current is less than speaker current, but so we dont load up the previous stage too much, we add more current gain with the driver transistors. This adds another 0.6V so the B of the top driver is 1.2V above the output & the B of the bottom driver is 1.2V below the output.
In total this gives a difference of 2.4V between B of the driver transistors.
The bias circuit is designed so there is a fixed voltage drop across it. We can adjust this so we can adjust the no load current through each output transistor, and in doing so we get to a point where the top & bottom transistors are both "slightly on", so we can reduce crossover distortion. If we didnt have this bias circuit, we woudl get nasty crossover distortion. As a very general rule, we adjust the bias so there is approx 50mA through each transistor.

Each power transistor has a power resistor in series with it. As there are slight differences between B-E voltage of transistors, this resistor helps force current sharing in each of the power transistors, and it also provides an easy place to measure the current in each transistor.

So for the LH500, Q308 is the bias transistor.
We know B-E voltage is approx 0.6V. The voltage across R319 (and part of the pot) is therefore 0.6V. If we assume Q308 base curretn is small, then the same current flows through R319 & also through R320 + D303 + D304. The volts across R320 is the current through R319 x 8.2k, + the voltage drop on the two diodes. When we adjust the trimpot, we are adjusting the curretn through R319 & therefore the current through R320, and what we end up with is the voltage across Q308 is approx the 2.4V we need to bias the output stage.

Even if you dont underatand exactly how the circuit works, it basically just provides a fixed voltage between the B of the driver transistors, so that we can get rid of crossover distortion. So knowing this, you could measure across C-E of Q308 & if you get approx 2.0 - 3.0 VDC you know this bit is working fine.

So after all that, the bias transistor does not generally blow up. Much more likely are power transistors, and often when they blow, they take out the driver transistors as well.
i have heard that Hartke tend to bias the amplifier "hot" so this means the power transistors will run warm even at no load. Also i have heard that there is not a lot of safety margin in their designs. You can run the bias lower, so the transistors do not get as hot, and the amp will be fine. I use 50mA per transistor as a general rule.

Also check R335 is also ok on each amp. This is part of a stabilisation network & this resistor must be good.

so, check power transistors & driver transistors and also the 0.22 ohm resistors, check the transistors mounting is ok & they are not shorting to the heatsink (if all power transistors are on one heatsink), check power resistors are not discoloured, check electrolytic caps dont have bulging cans, or have started to leak. Check any speaker protection circuits, especially if they have a triac in them.

Its preferable to check solid state amps WITHOUT a speaker connected, and it is ok to power up the amp without any power transistors installed.

Do the fans work ? unplug CN8, CN13, do the fans turn on, short out CN8 & CN13, do the fans turn on.
These amps need the fans or the power transistors will blow up.

So i hope this helps, now we can get back to fixing your amp.

[catstrat]

hartke lh 500

WOW. Thanks! That is a wealth of information. I'm always afraid because of hearing transistors are so fragile and i may do further damage by messing with it. This info definitely helps me out. I'll do some more checking and get back with everyone. Thanks.

[R.G.]

I'll add to the previous advice that most solid state amps will work, even if they sound horrible, if you simply short out the bias arrangements entirely, vastly underbiasing the output stage. This temporary un-fixing lets you find bad devices and other conditions without having to worry about thermal runaway and flaws in the biasing arrangements.

If it doesn't work with no class-AB bias, then something else is wrong. Once it runs with zero bias, you can start tinkering with proper bias.

[catstrat]

Once again, Thanks R.G. All of you guy's have been great.

[soundguruman]

I've confirmed it's in the output section and i have no schematic yet. I have a variac so i can raise the voltage to about 20 volts at the variac to do some voltage test but it gets hot. What voltages should i have at the base, emitters and collectors of the o/p transisters? I know it's not going to be full voltage but will i still have positive collector voltage at 1 trans bank and negative at the other. Thanks.

On this amp, pretty much everything from the output drivers and back is suspect.
The output drivers blow Q313, Q 314, and sequentially cause the (any or all) previous transistors to blow.
The output transistors do not blow as much, and "may" still be OK.
Anyhow if you work on it, replace the 2 output drivers regardless. I use a bigger rated part myself to replace them both.

The capacitors 10uF 50V C310, C311 should be replaced with 10uF 100V 105C, because those two tend to explode.
Replacing them with higher voltage and higher temperature rating seems to have solved the defect.
If they already exploded, then c what I mean.

On this amp, you should try to adjust the bias as far as you can, to cool it off.
It runs hot as stitska, even in idle conditions...very hot is normal for this amp, and that's why they don't last long.

In the meantime, take my advice, after you get it running...
check the thermistors, they go bad. These are mounted on the heat sinks.
One controls the fan speed.

MOVE the fan speed thermistor to the far end of the heatsink.
This is the part of the heatsink that gets hotter faster, and stays hotter longer.
MOVING the thermistor speeds the fan up faster, when any part of the heatsink is HOT.

The factory locates the fan thermistor in the wrong spot. The thermistor is mounted at the coldest place on the heatsink.
This prevents the fan from speeding up, even when the heatsink is damn hot (nuclear).
This probably contributes quite a bit to the short life span of the amplifier. It just runs too hot.

Also, you can tweak the fan circuit with a resistor, so that the fan is faster to start with.
This discourages heat build up...which is basically what is killing the amplifier.

A good way to approach this amp is...
just yank the circuit board and carefully test all the solid state components...
because a lot of them blow at the same time.
You don't want to miss a leaky transistor, etc...because it will just blow again.

Now this has a little circuit board tacked on to the main circuit board, which is a protection circuit.
If the output goes DC, or if the protection thermistor gets too hot, this little board shuts down the output relay.
But it does not shut off the power rails, burn baby burn.

[Enzo]

VOltage readings with the variac set at 20 volts are pretty much meaningless. The fact your current is already up to 4A means there is a heavy current draw in there. You either have shorted transistors, or your bias is open. Other things can occur, but those are most likely. Don't leave it cooking even at 20 volts. As soon as you see current ramping up, back off. Look for bad parts. There is no target variac voltage. You advance teh variac until either the thing is fully powered, or until current starts to ramp up. SOme dead short circuits run curren tup from the first volt. Other circuits can look pretty good until maybe 90v and then the problem reveals itself. Bottom line is that the meter to watch when variac-ing is the current, not the voltage.

Q308 is the bias transistor. You can tack a short across it from collector to emitter during tests. That will turn the bias cold, and if the ampo then is able to power up completely without trying, then the bias is the problem. Bias problems could be bad Q308, bad bias adjust trim pot, open diode or resistor. Oops, RG already said that in post #8.

Transistors can be checked in circuit for basics. If one measures shorted, pull it and check it by itself. There is nothing in any circuit that can make a shorted transistor appear to be not shorted. However circuits can make good transistors appear to be shorted, that is why you need to do the second test. But if anything checks as shorted, whether it is the transistor or a parallel circuit, that still will pop your fuses.

[catstrat]

hartke

All the info you guy's are giving me is going into my trouble shooting files for future reference. These are the troubleshooting tips that i'm just not familiar with until now. Thanks to everyone on this forum. However i had to put the Hartke to the side to handle an emergency concerning an Ampeg vt-40. Thanks.

[catstrat]

Hartke blowing fuses

What do you mean Mozwell when you say"non-zero" pertaining to testing the o/p transistors. I shorted across the bias trans from e to c and it still draws too much amperage. I removed all of the 5200 o/p transistors from the circuit ,leaving the 1943 transistors in, and powered it up to a full 115 volts with little or no amperage draw, less than 1/2 of an amp, if that. It looks like to me the bias circuit is ok. I removed all of the o/p transistors from the circuit and tested them with a meter. One of the 5200 shorted but the other one's i get nothing from the meter in either direction. I thought you will get a high resistance one direction and low resistance the other direction. I tested all the 2sa 1943 and none of them show a short. Thanks.

[soundguruman]

It's better to test em' all
better to be safe then sorry.

[Jazz P Bass]

removed all of the o/p transistors from the circuit and tested them with a meter. One of the 5200 shorted but the other one's i get nothing from the meter in either direction. I thought you will get a high resistance one direction and low resistance the other direction. I tested all the 2sa 1943 and none of them show a short. Thanks.

Well, it appears that the 'open' 5200's gave it all they had
The shorted one would definitely bring down the house of cards.