Triode Electronics VS Hammond

[airboss]

So I just picked up a Harmon Kardon 250 series integrated amp (Eflea) with a dead OT. I'm going to pull all the iron out of the transformer and see what it looks like at the inner windings. See if I can find an arc through the paper to the iron. I've been successful before in making repairs to the PRIMARY of OT, but never the secondary. Depending on what goes on the core first. Re-wax in a vacuum pot etc.

Anyway, I really like the user interface of the Epic amp, in other words it's got nice knobs and a nice faceplate. If I cant salvage the bad OT I'm going to use this thing anyway in my garage system after a complete strip and rebuild. It's got Baxendall tone controls and some other nice features. Some will change of course.

If it were you, would you go with the Triode Electronics A470 replacements (both of course) or the Hammond 1650H? I'm going to burn KT-66 instead of 6L6GB as in the original. The B+ is 380. I'm going to use Ultra Linear. Other transformers would be considered of course.

Thanks gang, fire away.
Dan

[dcgillespie]

Hey Danny -- In general, the Hammond transformers can't hold a candle to the Dynaclone transformers. The only thing in this case that favors the Hammond is that they all use a 40% UL tap, which is a very favorable point for 6L6 class tubes. But their power bandwidth and phase shift is not even in the same ball park as the Dynaclone. If you go the A-470 approach, you will definitely want to use fixed bias, and possibly think about regulated screen pentode operation as opposed to UL. For UL operation, 6L6 class tubes really need no less than 6600 ohms, where as the A-470 is 4300 ohms with 33% taps.

Dave

[airboss]

Originally Posted by dcgillespie

Hey Danny -- In general, the Hammond transformers can't hold a candle to the Dynaclone transformers. The only thing in this case that favors the Hammond is that they all use a 40% UL tap, which is a very favorable point for 6L6 class tubes. But their power bandwidth and phase shift is not even in the same ball park as the Dynaclone. If you go the A-470 approach, you will definitely want to use fixed bias, and possibly think about regulated screen pentode operation as opposed to UL. For UL operation, 6L6 class tubes really need no less than 6600 ohms, where as the A-470 is 4300 ohms with 33% taps.

Dave

I do like those tubes at 6600. Especially at 380 on the plates. So rock and a hard place. The question was asked knowing that Triode A470s are a superior transformer. I've quit using Hammond in my guitar amps and spec my own now.

So thanks, barring some radical shift that's what I'll do, regulate the screens, Dynaclones, and fixed bias. The HK is fixed, with it's current drive arrangement at -27.5 volts. They claimed 25 watts but it's really more like 18. Plenty for Altec VOTT drivers. I hope the transformer is a fixer. You got to pull all the Es and Is out of it, varnish, wax, etc to get at the interior of the coil. I'll bet it's an arc across the paper at the first layer. We'll see. I'm probably dreaming. It's unbelievably tedious. Forget it if the whole transformer has overheated and windings are shorted throughout. You'll never find out the interleave for rewind unless somebody finds a drawing from HK archives. This transformer has a primary to case short.

Don't tell anyone, but I've actually used PT with primary shorts with the case stuck up on a garolite board. Paint the top red and write "case hot!" on there. Not sure if an OT will cooperated with that scene.

Maybe somebody got a HK 250 OT eh? I'll bet a pair of Dynaclones would be cheaper than unearthing one of those. On the other hand, I could sell the iron off this project chassis to finance the Dynaclones. Maybe.

[DigitalJunkie]

One consideration is the extra heater current required for the KT-66 tubes,IIRC something like 1.25A each,versus the 0.9A for 6L6's. Be sure the power transformer can cope with that.

[Eli Duttman]

Danny,

Watch your step with KT66s, as you may overtax the power trafo filament winding(s). KT66s draw approx. 40 mA. more heater current, per tube, than 6L6s draw.

The SED (=C=) 6L6GC is a proven winner. I've been hearing good things about New Sensor's TungSol labeled, ST bottle, variant.

Edcor O/P "iron" is easily as good as Hammond and the price is lower. Edcor's model CXPP60-8-6.6K may be just what you need.

Tone controls come with their own set of problems. IMO, Max Robinson's "unity" gain implementation of the Baxendall setup is as good as it gets. To hold the bottle count down, change the AC coupled cathode followers to DC coupled ZVN0545A source followers. Be sure to have switches which totally remove the tone controls from the signal path, when they are not being used.

[airboss]

Well Gents, I've been wrestling with the pdf scheme to stick it up here so that we can all talk about what's available etc. My printer has decided it won't print pdf today. It did fine a few days ago, so I've messed up a driver default of some sort. And Adobe of course makes it so easy to change format on their prescious documents. I'm going to try a visio import/conversion or Word to change to .doc then go from their to another conversion to get to jpeg so I can post the dang thing.

I may have photoshop express around here somewhere, though I hate it, maybe I'll reload. Here's the url; pages 14-15 http://www.harmankardon.com/resource...picA250-om.pdf

Thanks

[Eli Duttman]

Danny,

Install Irfan View and then install Ghost Script, both which are free. Converting PDFs to other formats ceases to be problematic.

The DC heater supply is "interesting". Are you planning a revision?

The RCA jacks are probably spaced too closely for modern interconnect cables. That can be dealt with using a nibbling tool to create an opening and installing a piece of FR4 based and Copper clad PCB material therein. While retaining Faraday shielding, new RCA females can be spaced out to fit today's requirements.

Replacing the complicated source selection switching with something straight forward is (IMO) a good idea. Mouser stocks Lorlin switches, which are good performing and modest in cost. Check this catalog page out.

[wrenchone]

Eli, the tone control circuit looks interesting - I might do a partial rip-off when I get around to refurbishing (actually, redesigning) my old Lafayette amp.

[airboss]

Originally Posted by Eli Duttman

Danny,

Install Irfan View and then install Ghost Script, both which are free. Converting PDFs to other formats ceases to be problematic.

The DC heater supply is "interesting". Are you planning a revision?

The RCA jacks are probably spaced too closely for modern interconnect cables. That can be dealt with using a nibbling tool to create an opening and installing a piece of FR4 based and Copper clad PCB material therein. While retaining Faraday shielding, new RCA females can be spaced out to fit today's requirements.

Replacing the complicated source selection switching with something straight forward is (IMO) a good idea. Mouser stocks Lorlin switches, which are good performing and modest in cost. Check this catalog page out.

No way in the world that series DC supply stays in place but I don't want to waste the current. It will be regulated with ordinary of the shelf IC based regulator to 12VDC. I've been down that road before. With 123 VAC in the wall, that's a tube killer. One 12AX7 goes away. That's .33 amp at 6.3. I've seen these "scratch filter plates" and "Baxandall plate" before. It's a multi lead sealed unit. I don't know of anybody besides HK that used that stuff. OEM. So like I said, it won't be HK when I'm done, but it will be a handy size and look. I contacted a guy today that may have a transformer for me.

Yes to the switching changes. I dunno, it almost seems that by the end most of the stuff will be changed in some way. The wiring underneath has had many hands in there or HK had the worst QC in the world....something I find hard to believe.

[dcgillespie]

Danny --

A couple of things you might consider, even if you can replace the original OPT that is damaged:

1. This amp is a cathode biased amp. Converting to a true fixed bias system would get you a few more watts and lower the distortion all at the same time. Based on your comment that the amp puts out around 18 watts in stock form, that means that the original OPTs are probably around 6250 ohms plate to plate, which will work very well with 6L6 class tubes with fixed bias and the voltages offered in this amp.

2. Consider helping out the screens. Either some active regulation, or some zeners and a small choke to help limit the drop during increased power demands. Converting to fixed bias and bolstering the screen supply would do complete wonders for this amp. How are the OEM OPTs in the power bandwidth department? Can they handle any increased power at the frequency extremes that these changes would produce?

3. This amp has almost no NFB at all in the power amp stage. It's almost there for show only. With only a 12AU7 gain stage through the loss of a phase splitter stage and the relatively high 6L6 drive requirements, there is precious little OLG for the NFB loop to work with. Looks like they're trying to use some positive FB from the splitter to help the first stage out, but still, there's hardly any gain there at all. With the size of that FB resistor, I be surprised if there is even 10 db NFB. At a minimum, I'd consider a 7247 for the input/splitter tube, with the AX7 section as the gain stage. With the increased power that fixed bias and regulation could produce, the transformers may need some help in the NFB department if they are already stretched. At least it would help the damping factor.

All of this really depends on how good the OEM OPTs are, or what you end up replacing them with. Looks like there's lots of potential in this one for sure!

Dave