Replacing Farfisa Compact power transformer

[R.G.]

Looking at the power supply/reverb/preamp schematics the voltages don't make any sense whatsoever, especially SE-52. The one half of V2 is supposed to have 2.4V at the cathode with 1.8k cathode resistor (1.33mA) but that same current would have to do through the 220kohm plate resistor which would give 293V of drop off from a 245V rail, which doesn't make any sense to me.

Yeah - that's a real puzzler. I noticed that when thinking about the supply, and decided to put it down to either a schematic error or deliberate misinformation, which was and still is common. I just decided to go with the current calculated from the cathode. Something is wrong there.

And Ted's Antek power transformer is a real winner! Just wind on 3V to attach to one of the 6.3V windings and you're done, for $30 no less. Great catch, Ted.

[Helmholtz]

Looking at the power supply/reverb/preamp schematics the voltages don't make any sense whatsoever, especially SE-52. The one half of V2 is supposed to have 2.4V at the cathode with 1.8k cathode resistor (1.33mA) but that same current would have to do through the 220kohm plate resistor which would give 293V of drop off from a 245V rail, which doesn't make any sense to me.

There's no reason why the 2 sections of V2 should have different currents with the same value resistors.
So either the first triode voltages are wrong or it uses a higher value cathode resistor.

[glebert]

And Ted's Antek power transformer is a real winner! Just wind on 3V to attach to one of the 6.3V windings and you're done, for $30 no less. Great catch, Ted.

I ordered up the Antek, that seems like the best approach. Wondering if there is any reason to even generate 9VAC if I had a robust 9VDC rail to power the lamps instead. My thought was stack the 6.3V windings to get 12.6VAC -> modern rectifier bridge and caps (whatever that would give for DC, about 15V?) -> ballast resistor (ohms tbd) -> 9V regulator -> lamps and R504 (which drops voltage to 8V for all the cards).

[R.G.]

If all that 9V does is run lamps, yeah - run them on DC unless the circuit just won't do that. I tossed in an LM317 regulator because I view zeners as unsuitable for power applications in general. Some places, OK, but cheap, easy, high performing regulators are nearly always better. And the TO-220 package that most of them come in can dissipate 2W in free air without a heat sink. But if a resistor does it for you, why not. How much does the lamp current vary? Do they turn those things on and off, or leave them on?

Before just stacking the 6V windings, think a bit about heaters (do all/both tubes run on 12Vac, or do they need 6V?) and how the 6Vac gets grounded. The original grounded one end of the 6Vac. This is bad by modern practice. I guess if it works OK, you still can, but think about it. Also, if you full wave bridge the 12vac stack, then one end of that winding is one diode drop below circuit ground in pulses, whenever the bridge diodes are conducting. Ask me how I found this out... It can be a source of bad hum in some circuits.

Here's another thought. Full wave rectifying 6.3Vac gives about 7.5Vdc in a filter cap. If the lamps can run on DC and more or less stay on, maybe they'd work almost as well at 7.5V.

[g1]

Did you lift that HV winding and check the line draw (post #10)? I'd hate to see you go through all the work to swap the PT and find out C503 is leaky.

[glebert]

Did you lift that HV winding and check the line draw (post #10)? I'd hate to see you go through all the work to swap the PT and find out C503 is leaky.

I had forgotten to do that, so thanks for the reminder. Based on how much current it was drawing (glow on bulb) before I had the HV winding hooked up I didn't think it would make a difference and it really didn't, maybe it dropped the current down to 1.0A instead of 1.05A.

[R.G.]

I had forgotten to do that, so thanks for the reminder. Based on how much current it was drawing (glow on bulb) before I had the HV winding hooked up I didn't think it would make a difference and it really didn't, maybe it dropped the current down to 1.0A instead of 1.05A.

That's a little scary. A hundred watts is going somewhere. I wonder where. If the 8V output is really only a couple of hundred milliamps, how much is going out to the heaters and the 9Vac?

[R.G.]

Hmmm. I just had a thought. That equipment comes from an era when they often thought the AC wall socket provided 110Vac. Maybe its a victim of the modern move to higher AC lines, up at about 125 in my locale. It may be getting the transformer into soft saturation. It might be good to measure the AC current into the transformer with no secondary connected. If its over 50-100ma, that would be a clue about saturation.
Got a variac to see if primary-only current jumps when it's fed more than 112 or so?

[glebert]

That's a little scary. A hundred watts is going somewhere. I wonder where. If the 8V output is really only a couple of hundred milliamps, how much is going out to the heaters and the 9Vac?

Interesting development on this. I decided to pull the secondaries one at a time and see if the current dropped. Disconnected the 22VAC leads from the selenium bridge and the total current dropped from 1 amp down to about 125mA. I reconnected the HV to the silicon diode and caps and I'm getting 286V at the first node and 241V at the second. 120VAC current not even noticeably different with HV connected. So that was a little surprising. I am still not sure exactly how that 22VAC winding could be pulling that much current and still have the DC output be at 26V. The rectifier was mounted to the chassis and I didn't notice heat buildup there. From my reading on problems with selenium rectifiers the output voltage usuallly dropped (or it exploded into a toxic and awful smelling cloud). At this point I'm pretty much :shrug:

I guess next step is a new bridge rectifier and 9V regulator. Will be interested to see if it behaves.

If I end up not needing the Antek transformer I will certainly find something to do with it.

[R.G.]

Ah-hah!! 875ma of primary current to support a "22vac" winding.
From reading wikipedia on selenium rectifiers, they do degrade with usage but also un-form the rectifier junction inside in long storage. Good guess on replacing that with a silicon bridge.

[glebert]

I got a bridge that will fit decently into the small chassis so I hooked it up to the 22VAC windings and powered up on my variac and LBL. I didn't measure current but no glow on the bulb, with or without the neg side of the bridge output grounded. The peak voltage was at 35V so I need to get a higher rated cap for C502 (which had been replaced but only with a 35V cap).

I was planning on using a LM7809 regulator, which has a peak input voltage of 35V. Will a properly sized ballast resistor protect the regulator even on power on, etc? Regulator will mount to chassis for heat dissipation.

[glebert]

I installed the new smoothing cap for the "22VAC" bridge output, 9V regulator, changed the ballast resistor from 75 ohm to 120 because the higher voltage out of the silicon bridge. Powered it up and organ is working and drawing 155mA, or pretty much 19 watts exactly.

Moral of the story, if there is selenium rectifier anywhere near a problem or suspected problem, replace it!

Thanks for all the advice and help!

[Jazz P Bass]

Nice outcome.