Harman Kardon A-30K (A-300) power supply

[depaj]

Hi!

I just picked up a Harman Kardon A-30K, the kit version of the A-300. It works but someone replaced the power transformer from a 110V primary to a 250V but in doing so he also replaced the doubler circuit with a classic half wave rectifier fed from a 250-0-250v secondary winding. Only problem is that we only have 230V mains here so the secondary is only putting out 210V giving me about 290V rectified and only 270V under load when the amp needs about 350V to work properly.
Now my question is what would be the best option :
-Reinstate the doubler circuit and find a power transformer with the original specs (as they're not indicated on the schematic, does anyone have these ??)
-Find a different power transformer with the correct primary and secondaries to keep the actual rectifier circuit ?

Thanks !!

 

[gadget73]

What does your heater voltage measure at? if that is also low, it may actually make sense to use an external boost transformer to bump the voltage coming in to the amp. That might get it close enough to work.

another possibility is converting it to a bridge rectifier. That should also bump the voltage.

 

[Tom Bavis]

Bridge would give over 550V - not really an option. I'd be inclined to live with the 10 watts that this would deliver instead of 15 (as long as the transformer is adequate - doesn't overheat).

 

[depaj]

What does your heater voltage measure at? if that is also low, it may actually make sense to use an external boost transformer to bump the voltage coming in to the amp. That might get it close enough to work.

another possibility is converting it to a bridge rectifier. That should also bump the voltage.

Heater voltage is 6.2V AC. I tried wiring the transformer for 220V (it only has this other tap, no 230 or 240V) but that bumped the heater voltage to 7.8v so no good.
As Tom Bavis said, converting it to a bridge rectifier would be way to much and getting it back down would dissipate way too much heat.

Bridge would give over 550V - not really an option. I'd be inclined to live with the 10 watts that this would deliver instead of 15 (as long as the transformer is adequate - doesn't overheat).

The transformer seems rugged enough, it's just sad not to have this little amp working at its full potential. Also it looks really bad the way it has been done.

 

[gadget73]

Bridge would give over 550V - not really an option. I'd be inclined to live with the 10 watts that this would deliver instead of 15 (as long as the transformer is adequate - doesn't overheat).

with a cap input, true. Choke input would be pretty close though. Whacking a couple things in PSUD real quick comes out to 370 volts.

 

[TriodeLuvr]

Your best option is to replace the transformer with one that's correct for the task. There are companies in Europe that sell 230V transformers for guitar amplifiers that use 6V6 and EL84 outputs. This will be much closer to duplicating the original conditions. If you're not familiar with those companies, you might consider posing the question on a guitar amp forum. The site below has a number of European members who build their own amps and who have discussed their European sources.

https://www.18watt.com

Jack

 

[depaj]

with a cap input, true. Choke input would be pretty close though. Whacking a couple things in PSUD real quick comes out to 370 volts.

That would actually be pretty close, I could add an extra resistor because one of the caps won't be used for the doubler circuit. Any suggestions as to what the inductance and resistance should be ?

 

[depaj]

Your best option is to replace the transformer with one that's correct for the task. There are companies in Europe that sell 230V transformers for guitar amplifiers that use 6V6 and EL84 outputs. This will be much closer to duplicating the original conditions. If you're not familiar with those companies, you might consider posing the question on a guitar amp forum. The site below has a number of European members who build their own amps and who have discussed their European sources.

https://www.18watt.com

Jack

That was my first idea but it's probably the more expensive one. Also would you suggest keeping the voltage doubler circuit as originally designed by HK (and in that case what would the secondary voltage need to be ?) or use a transformer with a higher secondary and keep the rectifier as is ?

 

[gadget73]

Making it as close to stock with just a higher primary voltage in a transformer that physically matches the original would be the cleanest solution. Doubler secondary for 350v probably in the ballpark of 160 vac but psud2 would be a better model if there are no values on the schematics

 

[Tom Bavis]

The voltage doubler or a full-wave bridge will best, as they make most efficient of the transformer. The center-tap full-wave circuit needs a transformer about 30-40% larger for the same DC output. That's why I wondered if the replacement was adequately rated. Is there a cover? Will limit the transformer size.

 

[TriodeLuvr]

That was my first idea but it's probably the more expensive one. Also would you suggest keeping the voltage doubler circuit as originally designed by HK (and in that case what would the secondary voltage need to be ?) or use a transformer with a higher secondary and keep the rectifier as is ?

The vast majority of guitar amplifiers use a center-tapped full wave power supply, the same as what you have now. This is almost certain to be the least expensive option considering your location in France. Regarding the choke input filter that's been suggested, the reduced output voltage (compared to capacitor input) relies on a certain minimum load current. The fact that current draw by the tubes is non-existent until they warm up and conduct means you must either A) rate everything in the amplifier to withstand a significant voltage surge at turn-on or B) install bleeder resistors sufficient to load the supply and maintain critical inductance during warm-up. These resistors can consume a considerable amount of power relative to a capacitor input supply that doesn't need them.

Jack

 

[depaj]

The choke input seems like a good idea but I'm not really keen on modifying the whole thing and there really isn't that much room for a choke underneath. There is a cover that limits height but I'm not sure that it will be on there very often. I would really like to keep it as original as possible so maybe going back to the voltage doubler circuit Is what I'll go for. The schematic only indicates dc voltages so I don't have the original specs for the transformer. I'll probably give PSUD2 a go but if anyone has one of these amps and can read or measure the output of the transformer that would be even better .

 

[TriodeLuvr]

I'm a big fan of the various HK series of tube amps. One of the models in my collection is an unrestored TA3000X that's been put away in my garage the last 20 years. This has the same output section as the A300, so probably the same transformer voltage. Unfortunately, my test bench is completely covered with a large DIY monoblock SET project at the moment. I just can't clear it off for a test right now.

SAMS Photofact always lists replacement transformer voltages in their service manuals. Unfortunately, following an extensive online search, I have come to the conclusion that SAMS never published the A300 or TA3000X. And as I'm sure you know, the factory HK manuals and schematics don't show transformer voltages.

All this aside, I can tell you that the secondary voltage of a doubler transformer for the B+ of your amp will be about 150-160 VAC. Because of the high current draw of the doubler, the DC resistance of the secondary winding should be only a few ohms. These requirements aren't typical of more modern tube gear, so it's very difficult or impossible to find general replacement transformers suitable for doubler service. This - plus your location in France and the need for a 230 VAC primary - is the reason I suggested buying a center-tap transformer from a European source.

After a little more thought, I realize some guitar amplifiers have also used a FW bridge configuration. The Peavey Classic 30 is one example. It uses (4) EL84 in the output stage with a B+ of about +330 VDC. As noted by another member above, the FW bridge would be more efficient than the center-tap design. My point is that European sources of power transformers made for the guitar industry will probably be the most practical option for a transformer that is electrically and physically suitable. Again, this is worth discussing on the forum I mentioned (or other guitar-related sites) before spending too much time searching for an exact replacement that may not even exist.

Jack

 

[depaj]

I've been away a few days and didn't have time to do extra research until today. It seems indeed very hard to find anything coming even close to the requirements necessary for the original doubler circuit. So for the use of a FW bridge I would be looking for a 0-265V and a 6.3v secondary right (I did a simulation with PSUD2 but don't really know what load resistance I should enter, also the actual transformers secondary dc resistance is about 200ohms which seems high, the simulator is set to 31ohms, I guess this would be closer to what I'd get with a modern transformer) ?

 

[gadget73]

if there are any current values on the schematic, use those with a constant current load if you can

as for the transformer, if you know the DC resistance values, use the calculator in psud to figure out the transformer impedance. On the box where it lets you put that in, there is a second button that will take you to the calculator. In there you put in the various voltages and resistance values and it calculates it for you.

 

[depaj]

There are no current values on the schematic. The first resistor has about 32mA passing through and by calculating the voltage drop across the output transformer winding from centre tap to plate divided by the measured dc resistance I get about 52mA per tube so that would add up to about 240mA total.

I actually wanted to calculate what the needs would have to be for the replacement transformer so the parameters of the one in place won't help much.
If I consider B+ to be exactly 355V and assume the amp is drawing 250mA can I simply calculate the load impedance as being 355/0.240=1.5Kohms ?

This leaves me with only the impedance of the replacement transformer as an unknown variable that makes quite a lot of difference in the voltage requirements...

If I leave the transformer impedance at 31ohms, I need about 280V output.

 

[gadget73]

you can do it that way, or just set a current load at 250ma, comes out to the same place really. The one advantage of the current load is you can step it from idle current to full output current if you really want to model the supply performance.

 

[GordonW]

The voltage doubler or a full-wave bridge will best, as they make most efficient of the transformer. The center-tap full-wave circuit needs a transformer about 30-40% larger for the same DC output. That's why I wondered if the replacement was adequately rated. Is there a cover? Will limit the transformer size.

Tom is getting at something important here. The A30K uses a voltage doubler, which like a bridge rectified supply, will act like a lower impedance power supply, given the same rating power transformer, than a standard center-tap full wave supply.

I would want a transformer that could either do bridge or voltage doubler in this amp. You will hear it, in improved bass output. These amps can do some pretty astonishingly good things in the bottom end, when the power supply is right.

Regards,
Gordon.

 

[gadget73]

on the topic of that, the stepped load is where you'll really see the difference in supply performance. The less droop the supply has, the tighter the bottom end tends to be, or at least thats my experience with tube amps.

 

[GordonW]

on the topic of that, the stepped load is where you'll really see the difference in supply performance. The less droop the supply has, the tighter the bottom end tends to be, or at least thats my experience with tube amps.

Yeah, that's usually the case. That usually correlates.

Regards,
Gordon.