• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

Heathkit AA-100 iron with 6L6GC

thorpej

thorpej

AK Subscriber
Subscriber
I recently nabbed a non-working Heathkit AA-100 for a good price... seller said "transformers guaranteed", and I just finished testing them and they're perfect.

I'm thinking of building another variant of my Mullard 5-20 / Eico HF-89-inspired 6L6GC amp using the power and output transformers. I drew up a quick load line using the GE 6L6GC plate characteristics graphs and at 450V plate, 250V screen, looks like I can manage 28W before transformer losses...

But note how the class B line goes right through the knee of the +4V grid line... Inferring from the transfer characteristics graph, I think I can nudge the 0V grid line up to that spot by doing 300V on the screen. That would get me to 40W before transformer losses. Seems like I could manage a decent 35Wx2 amp with this configuration.

Sanity check, please? Thanks!
 

Attachments

Register to hide this ad
Are those AA-100's rated for 40 watts at 20 Hz? I have a pair meeself, and I'm not sure..they don't look that big.

If you use the "average transfer characteristics" graph from the GE 6L6 data sheet, you should be able to eyeball fairly accurately where you'd need to put the screen volts to get the knee of the 0V curve to cross your load line. (i.e., it may be easier to interpolate the average transfer characteristics graph than it is the plate voltage graph.)

Just quickly eyeballing the available graphs from the March 1959 GE 6L6-GC data sheet, looks like the top graph on page 8 is the graph you'd need. A quick glance says 300V screen might be just a tad too much.
 
Yah, ratings wise, not really sure. The original specs for the AA-100 were 25Wx2 30Hz-15KHz ..but there could be other shortcomings of the circuit that limited the response.

Ah, thanks for pointing out the graph on page 8 ... I was using the top graph on page 5 and extrapolating... Yah, you're right, 300V looks like too much... looks like 275V is closer to what I want, but maybe still not quite right. I'm planning to build a simple regulator for the screen voltage, and will use a trim pot to be able to dial it in as needed.
 
I dunno. From some sims I've done, I'd bet you're probably pretty close with 300v on the screens. That, with about -26 to -27v bias, should be great. Without transformer losses, it should theoretically be able to make about 38 to 39 watts- assuming a max of about 20% in transformer losses (IIRC, that's what Dave Gillespie mentioned somewhere?), it should be able to do 30 watts, in the midband. There is, as was mentioned above, some question about whether the AA100 transformers will do much more than 25 watts at LF, but that's something that would only be known, once the amp is running....that said, the lower effective plate impedance to be had with higher screen voltage, would help a bit in countering low primary inductance issues (LF saturation).

I'd definitely make it adjustable between 275-310v or so on the screens...

Regards,
Gordon.
 
thorpej said:
For the screen regulator, I was planning to use the extremely simple Leigh Bassett circuit (A Solid-State Filter Choke or Field Coil Replacement), really easy to make adjustable within a given range. Then just use large enough screen stoppers so that the voltage gets pulled down if the screen current starts to run away.
Click to expand...

I don't know if the original power transformer will necessarily support the additional current required for higher output,regardless of which power tubes you use.
IIrc,the wattage spec of this unit was met ''single channel driven'',and was substantially less in stereo,as usual probably due to power supply limitations.On a positive note,I have an AA-100 also,and the output iron seems reasonably substantial.

The circuit you posted the link to looks like a fairly simple but very effective design.But when I saw the regulator number,I thought ''good luck finding that,and if you do,how much is it gonna cost?'' Hmm,readily available, for slightly less than peanuts....go figure...

Seems like whenever I find a circuit I want to try,whatever semiconductos they use are long discontinued and only available for the price of a second mortgage:(
 
Buy a Hakko soldering station. Bespeaks quality. Easily adjustable. Why buy a used/non functioning antique? Why make things difficult for yourself? YMMV.
 
arts said:
I don't know if the original power transformer will necessarily support the additional current required for higher output,regardless of which power tubes you use.
IIrc,the wattage spec of this unit was met ''single channel driven'',and was substantially less in stereo,as usual probably due to power supply limitations.
Click to expand...

I was kind of wondering about that. The Photofact for the AA-100 quotes 140mA, which is pretty awful -- barely enough when the amp is idling, even with the original 7591s. OTOH, I will be using SS rectification which should provide a significant improvement to the impedance of the power supply in general.
 
thorpej said:
I was kind of wondering about that. The Photofact for the AA-100 quotes 140mA, which is pretty awful -- barely enough when the amp is idling, even with the original 7591s. OTOH, I will be using SS rectification which should provide a significant improvement to the impedance of the power supply in general.
Click to expand...

The original 5AR4 rectifier is about as good as it ever got in consumer types,and only loses about 15-20 volts at full rated current.The SS rectifiers will certainly raise the voltage (and unfortunately may require higher rated or series'd filter caps) and lower the supply impedance,but that won't help much if the transformer sags like a pair of old boob's under heavy loads.

Losing the 5v@2a (10w) rectifier heater load may help a bit,but you're still constrained by the wire size/resistance of the hv secondary.

Regardless of relevance,big +1 on the Hakko soldering station:)
 
arts said:
The original 5AR4 rectifier is about as good as it ever got in consumer types,and only loses about 15-20 volts at full rated current.The SS rectifiers will certainly raise the voltage (and unfortunately may require higher rated or series'd filter caps) and lower the supply impedance,but that won't help much if the transformer sags like a pair of old boob's under heavy loads.

Losing the 5v@2a (10w) rectifier heater load may help a bit,but you're still constrained by the wire size/resistance of the hv secondary.
Click to expand...

The Photofact shows the resistance as 53 ohms from the CT to either end of the HV winding, and claims 740VCT @ 140mA. I suppose I should go test the unloaded voltage again (I forgot to write it down yesterday), put 106 ohms as the series resistance of the voltage source in Spice, and make sure I am getting what I expect with a simulated 140mA load.
 
You may find that it's easier to deal with the B+ voltage with that power transformer and SS rectification, if you use the 5v heater winding in series with the primary winding, to reduce the secondary voltage. You should be able to get about 355VAC per side, that way (instead of 370V). That gets unloaded B+ below 500v- which makes caps easier. Loaded down, you should still be able to get close to 440-450v, I'd think...

Regards,
Gordon.
 
Ok, measured the resistance on the HV winding ... 95 ohms for the entire winding (~47 from each end to the center tap). With 120VAC on the primary, the unloaded secondary is 402-0-402. If I punch those into the simulation with a single 330uF filter cap followed by a 140mA current sink, I end up with 503VDC +/- 1VDC ripple. 503V seems a tad on the low side, but isn't too far off.

I suppose another option, to avoid over-taxing the PT in terms of current draw, would be do something based on 7189 / EL84M. Of course, the "problem" is that I have a stash of 6П3С-Es and no EL84Ms :-)
 
I've built a couple EL84 based amps with AA-100 iron. What I did was used a R-C input filter to tame the B+, around 50r to 100r, ended up being R-C-L-C. The rest was a typical circuit similar to a Dyna ST-35.

They sounded really good too!
 
Seems like you have a workable power supply, but given the max current rating on the power transformer, output power might be a bit less than your spec'd 35W, both channels driven. You might be down in the upper 20's to low 30's range. Still decent though, I think.
 
Yah, and if I keep the screen regulator adjustable over a wide range, then it should be relatively straight forward to dial in the operating point.

In pentode mode, I won't need as much voltage swing to drive the output stage, so I might dip into my 6SL7 stash for the AF amp stage. :thumbsup:
 
Hi Thorpej, about your project i suggest you use the 6L6G tube. You would have no worries about your power supply and only would have a little lower power output in the end. But the sight of those 6L6G would be like baby WE 300s with their sexy curves and all. You know nature abhors a straight line! Oh, and 6SL7 + 6L6 a match made in tube heaven. cheers.
 
I've been mulling this over and I think I've decided to go a slightly different direction and go with a 7189 / EL84M power section instead. It's a better fit for those output transformers anyway. Use of an EFB(tm) cathode regulator instead of traditional fixed bias will allow me to get the voltage across the tube in the right spot (Va-k would have been too high otherwise). And I will still plan to use a regulator to get 300Vg2-k. Now to decide on a front-end topology ... I don't need nearly as much signal swing to drive EL84Ms, so a 12AX7 AF amp stage direct-coupled to a 12AU7 cathodyne phase inverter is probably a better fit than the Mullard 5-20-like front-end.
 
You must log in or register to reply here.
Back
Top Bottom