DHT OTL Linestage - Tram 2

[decky]

Originally Posted by ThorstenL

Hi,



It is a bit much, try a 0.33 Ohm / 9W resistor in series with the heater winding.

Ciao T

I am a bit confused. Nominal secondaries are 5.5VAC and mine are 6VAC . Are you saying that 0.5V difference would cause a meltdown of the regulator? It looks to me that heatsinks are simply not designed properly. They are the biggest that could fit in a nice box but not the ones that can do the work.

What happens with people who buy the amp assembled and expect a final product without wanting to modify anything?

[ThorstenL]

Hi,

Originally Posted by decky

I am a bit confused. Nominal secondaries are 5.5VAC and mine are 6VAC . Are you saying that 0.5V difference would cause a meltdown of the regulator?

I am not saying it will lead to the regulator melting down. As you note, it can run quite a bit hotter than what it does. Adding the resistor will take up some of the extra heat.

Originally Posted by decky

What happens with people who buy the amp assembled and expect a final product without wanting to modify anything?

If we are told that the customer has an abnormally high mains voltage we will fit these resistors ourselves. Details for this are actually in the Filament supply manual, I need to check if they are in the Tram Manual and if necessary update that.

Ciao T

[decky]

Here is my calculation and you tell me if I am wrong. Under assumption that you have the series resistor in place and that regulator current to the ground is not more then 100mA dissipated power is P=(Vin-Vout)*ILoad+Vin*Ig where input voltage is 8.46V with 6VAC supply, Vout is 2.5V ILoad is ~1.88A yields with 12W of heat. If you assume the ambient temperature of 30C junction-case resistance of 2.7C/W, case - heatsink resistance of 2 and heatsink resistance of 8 - regulator junction temperature is 170degC.

Did I make a mistake somewhere?

[ThorstenL]

Hi,

Originally Posted by decky

Here is my calculation and you tell me if I am wrong.

The RMS voltage before the regulator is normally around 6V (you need to account for losses in the diodes and the ripple which at 2.5A DC is quite substantial), based on actual measurement, so it drops around 3.5V (a little less in reality due to some additional circuitry) with 2.5A Current, so real dissipation is around 8.5W.

The heatsink I am not sure of the exact K/W, but based on my measurements of it's temperature I'd put it at 6K/W. The chip is around 2.5K/W and the insulation normally (correct assembly given) around 0.5K/W.

So total thermal resistance is 9K/W and 8.5W will raise the chip junction to around 75 Degrees above ambient, which with the Tram II is likely around 35 Degrees. So Chip temperature would be expected around 110 Degrees, with some headroom (not much, admittedly) for hot temperatures and mains over-voltage.

This also matches my tests on the actual supplies.

Ciao T

[decky]

Originally Posted by Desmo

@Thorsten: Thanks for helping out with answers...! I have now recieved my Tram 2 and I have a few issues that I hope you can help me with...

1). C16 on the autobias module blows up! I don't have a circuit diagram, so I have no idea what this 100uF, 25V elco is doing, but after a few days it blev up (metal housing was shot of the cap). I hoped it was just a bad cap, so I have replaced it but the new one is beginning to ''lift it's top'' as you can see in one of the attached close up pics. Can you help me with a solution to this..!?

2) HEAT... I run 2a3 tubes and the heat sinks on the DHT supply modules get so hot, that the supplies are getting unstable after +5 hour. Unstable = noise heard through the speakers and low frequency pulsing (woofers flapping) that indicate some sort of low frequency oscillation. I should mention, that I have the preamp located on an open shelf with lot's of ventilation around it. Also worth mentioning is that I have tried with both 0.47 ohm resistors and 0.22 ohm resistors on the input of the DHT heater supply modules, but the modules can't regulate the output to 2,5V dc with the resistors installed.

I have tried a ''proof of concept'' approach to increase the cooling area of the heat sinks, see attached picture, and with the added surface area on the heat sinks the modules does not get unstable, however they still get VERY HOT..!

Hi Desmo - you are not alone - mine bias module had the same failure as yours just on the other side of the PCB after less then 3 hours of operation. Your cap looks like a fake nichicon - mine was a cheep Chinese one. All of them are replaced now with Panasonic FC series.

Also the same problems with the heater supplies - as you have probably read already.

This pre-amp has soooo much potential it is a pity to loose control over such small details that can render the whole amp completely useless

[decky]

Originally Posted by ThorstenL

Hi,



The RMS voltage before the regulator is normally around 6V (you need to account for losses in the diodes and the ripple which at 2.5A DC is quite substantial), based on actual measurement, so it drops around 3.5V (a little less in reality due to some additional circuitry) with 2.5A Current, so real dissipation is around 8.5W.

The heatsink I am not sure of the exact K/W, but based on my measurements of it's temperature I'd put it at 6K/W. The chip is around 2.5K/W and the insulation normally (correct assembly given) around 0.5K/W.

So total thermal resistance is 9K/W and 8.5W will raise the chip junction to around 75 Degrees above ambient, which with the Tram II is likely around 35 Degrees. So Chip temperature would be expected around 110 Degrees, with some headroom (not much, admittedly) for hot temperatures and mains over-voltage.

This also matches my tests on the actual supplies.

Ciao T

Thanks for clarification but I cannot accept 6K/W for the heatsinks - I have some To220 heatsinks rated at 7.5K/W and they are 30% bigger then what you use and they cannot fit in the box . Also with the temperature rise taken into account and efficiency drop you are looking at 10-12K/W if you are lucky even if with only 30deg rise over ambient. I was very optimistic with my estimate of 8K/W. I agree with you for 0.5K/W case to sink - I put some arctic silver and thin mica film there - that should help.

So all in all we agree they are borderline acceptable - I might try running 45 valves with extra heatsinks epoxied on the exiting ones - similar to what Desmo did. That might solve the problem .

What do you think about putting the heater supply in a separate box with much bigger sink. Do you think that extra cable length you influence noise pickup significantly?

[ThorstenL]

Hi,

Originally Posted by decky

Thanks for clarification but I cannot accept 6K/W for the heatsinks - I have some To220 heatsinks rated at 7.5K/W and they are 30% bigger then what you use and they cannot fit in the box .

My 6K/W come from actually measuring the heatsink temperatures, metal tab temperatures and the dissipation and then working backwards from these, not a manufacturers rating.

Please add the resistors, you should be able to get heatsink temps in the 80 degree range with 2A3's.

Ciao T

PS, I am investigating the popped Capacitors. It appears the silk print on the PCB is inverted. What worries me is why these cap's, despite being reversed, have not been blowing up in ever so many Amplifiers in use for years and they only go in the Tram and only some times.

I'll confirm the polarity later, but if you have C16 going again, try it reversed.

BTW, the Cap does not much where it is anyway (added in a "belts and braces" fashion, it can also be safely left off.

[decky]

OK no problems

I will try with the R33 resistor, extra cooling and a friend has a 240 to 220 step down trafo that will help as well.

Keep me posted about that C16 - the same cap died in mine - it looked like the other side of the board in the pic but it is the same cap.

[Desmo]

Originally Posted by decky

Hi Desmo - you are not alone - mine bias module had the same failure as yours just on the other side of the PCB after less then 3 hours of operation. Your cap looks like a fake nichicon - mine was a cheep Chinese one. All of them are replaced now with Panasonic FC series.

Also the same problems with the heater supplies - as you have probably read already.

This pre-amp has soooo much potential it is a pity to loose control over such small details that can render the whole amp completely useless

Good to know that others also has the same problems (meaning that when one of us finds a solution, we all have a solution)

Regarding the heat issue on the DHT supply modules I tried first with the supplied 0.47 ohm resistors in series on the AC line feeding the modules, but the voltage drop was too much for the modules to regulate. I simply could not get a 2,5V DC on the output with 2a3 tubes. Then I tried 0,22 ohm resistors, but the highest DC voltage I could get on the output was something like 2,2V DC.

I then dropped the idea with series resistors and instead I tried adding more cooling area to the heat sinks as you can see in a previous post. With the added cooling area the modules no longer get's unstable because of overheating (no more woofers flapping), but the chips/heat sinks are still extremely hot. My plan is to install larger heat sinks AND to improve the ventilation! Air can get in trough the bottom, but there is no way for it to escape through the sides or top. I think this will be needed to get a more ''permanent fix'' to this issue.

Regarding C16 I had this reply from DIY hifi supply (through Peter at Acoustic Dimension in Holland where I bought my Tram). It's a bit difficult to understand from the message, if the cap should be reversed to the opposite of what's indicated on the silk screen? Or if he is talking about the need for a cap with higher voltage rating? Or both?

Hello Peter,

is the cap on C16 position? we just discovered that the +ve and -ve of the cap is inverted in the soldering. The problem doesn't show up in other application but only tram. We just discovered it. Seems like when the current of the tube is high then overheat the wrong cap.

regards,
Simon

[ThorstenL]

Hi,

Originally Posted by Desmo

Then I tried 0,22 ohm resistors, but the highest DC voltage I could get on the output was something like 2,2V DC.

Hmmm, still sounds like something wrong.

Originally Posted by Desmo

Air can get in trough the bottom, but there is no way for it to escape through the sides or top.

Actually, air should be able to escape quite well. The tube panel is mounted with qute a gap and Air should exit around the tube sockets.

The heatsinks should be at around 80 degrees in operation.

Originally Posted by Desmo

Regarding C16 I had this reply from DIY hifi supply (through Peter at Acoustic Dimension in Holland where I bought my Tram). It's a bit difficult to understand from the message, if the cap should be reversed to the opposite of what's indicated on the silk screen? Or if he is talking about the need for a cap with higher voltage rating? Or both?

The printing on the PCB that shows the polarity of the capacitor is wrong, somehow, somewhere the print for this capacitor was flipped, so if it is put in as shown, it will be in the wrong polarity.

Funnily enough, this only seems to cause issues in the Tram and then only in a small number. It is no issue in any number of Amplifiers, inclusing one I have been running for more or less 3 Years every day...

Ciao T