Shindo schematics?

[Peter Daniel]

Beware, Silver Mica are in my experience the capacitors closest to a piece of wire... This can be a bad thing.

That could be a good thing if your stereo is built from a piece of wire, unfortunately it's not.

Silver mica from Sayal is not worth even trying. I like to quote that post when it comes to silver mica, as it's very close to how I feel about those caps: http://www.diyaudio.com/forums/showthread.php?postid=262664#post262664

 

[ThorstenL]

Konnichiwa,

Silver mica from Sayal is not worth even trying.

I would advise to try it anyway.

I like to quote that post when it comes to silver mica, as it's very close to how I feel about those caps: http://www.diyaudio.com/forums/showthread.php?postid=262664#post262664

As much as I respect Jonathan (we had lengthy and friendly exchanges over at the joe list in the elder millenias and I learned a lot) he writes in the context of all solid state direct coupled, high feedback circuitry (as incidentally do you), while I write from the context of transformer coupled output usually low/no feedback circuitry, though my testing was done in a wire bypass test setup, so are about absolute neutrality, not about what happens in a given circuit.

So I suspect the differences in opinion and experience are quite explainable.

Sayonara

 

[Peter Daniel]

That actually may be the explanation. I also know that some of those silver micas are much better quality that what's available here. Something like that maybe: http://www.diyaudio.com/forums/showthread.php?postid=22788#post22788

 

[rdf]

First, what opertaing points do you use? Especially the E180F can sound bright and steely if the anode current is too high, try 6mA or less, given that you probably have a low anode voltage as well.

Thanks for that! First though I should clear up a "posting-from-work" ommission. That should have read an E180F grid-leak biased with a TRW in front, the leak resistors 10 Meg Caddock spares from an earlier project. To my ear this has so far provided the cleanest top end performance. The plate load (today) is an IXYS chip surounded with stopper resistors and with 10-turn trim pot to facilitate easy bias current changes live. Prior to that it was VR90 regulator driving 150H Hammond chokes as plate loads, which sounded very good. They're only rated for 8 ma however so the IXYS's went back in to try something higher. The current right now is in the 12 ma range with around 110 volts on the plate. With the chokes it was between 7 and 8 depending on tube and 80-90 on the plates.

Regarding the EL84 op point it might be easier to say what wasn't tried. If it was between 8 and 12 watts standing dissipation and between 130 and 270 ohms in the cathode, it did time. Tubes include EH, Sovtek, JJ and the Anthony Welsh Russkie jobs. I keep gravitating back to about 255 volts plate-to-cathode with around 40 ma plate current (180 ohms) more on feel than anything else.

Secondly, both E180F & EL84 are very high transconductance, stick 1K gridstoppers on anything that looks like a grid (screen, suppressor etc.), even if you triode wire the valve. Also, if you triode wire remember to take any freely available suppressor grid to the anode, rather than the cathode.

Of those suggestions only the EL84 is without gridstopper. Already did time with the front end pseudo triode, very impressive frequency response (flat to 120 kHz into load if I recall) but again sounded better to me in true triode. The scope shows both plates clean as a whistle to 60 MHz at least. I'll revisit the stopper for the output and intend to play around with variations of g2 and g3 anode strapping on the E180F. However this might be another one of those things which on revisit makes me wonder what I could have been thinking the first time.

Thirdly, go for fixed bias on the output stage as well and as the EL84 has loads of gain and transconductance apply cathode feedback from the output transformer secondary (observe correct polarity) to the cathode of the EL84, in other words connect the EL84 cathode to the output transformer secondary so that you attain negative feedback.

That is interesting, I had considered that form of cathode feedback but not fixed bias. I'll give it a shot. Certainly the E180F won't break a sweat over the extra drive. Since I have these Hammond 156C 150H chokes laying around, what do you think of the following for a grid supply?

rect - cap - IXYS CCS - 156C - resistor in parallel with cap - grid resistor

The current through the active CCS sets the voltage across the final resistor, which doesn't need to be any higher than 1500 ohm to arrive at a reasonable grid bias voltage using a reasonable current of 8 ma or less. The 156C creates an very benign load for the chip as well. Should be very quiet if nothing else.

Forth, the EL84 is extremely eay to drive, try an unbypassed cathode resistor to bias the E180F, if you get too much top end rolloff add a small value silver mica "tuning capacitor" across this unbypassed cathode resistor to get any drooping trebel response back up.

I tried the E180F with cathode degeneration before converting to gridleak. The gain went up on removing the bypass cap, which I take to be an indicator of Bad Things. It's something I meant to revisit but converted to gridleak. It seems to really like having its cathode hard grounded. Save for the usually SE low frequency stuff the amp's hiss is at the limits of audibility with an ear pressed to the dust cap of a 12" Tannoy Gold.

Fifth, remember the that output stage current loop is completed through the powersupply. The final powersupply capacitor matters a lot sonically. You can add a modest value (22..47R) series resistor after the final electrolytic capacitor with a largish motor run film capacitor (22uF or larger) to ground as effectively "decoupled bypass" capacitor.

Right now the two EL84's share two 60uF and one 40 uf ASC motor runs (for a total of 160 uF) with 6 uF Solen fastcap bypasses. I don't like the Solens in a coupling position but they appeared to help here. There's also a shipment of 100 uF Japanese ASC's on the way. No electros anywhere to be seen, even the first cap of the LCLC supply is a 30 uF Solen I had laying around. What capacitance would you consider 'enough'. I also have a pile of supply chokes around and may 'Y' off to right and left after the first LC section.

You might find that with only one coupling capacitor and one PSU capacitor plus local feedback in both stages for better linearity will make this a wholly different animal.

Let's hope! Since the output stage appears to be the main limitation at this point - the front end measures more linear by a factor of 20 - I'll start there.

Beware, Silver Mica are in my experience the capacitors closest to a piece of wire... This can be a bad thing.
[/B]

Dismaying, yes. Bad, never. The iron's heating up right now. Thanks again for your input, very greatly appreciated.

 

[ThorstenL]

Konnichiwa,

an E180F grid-leak biased with a TRW in front, the leak resistors 10 Meg Caddock spares from an earlier project. To my ear this has so far provided the cleanest top end performance.

Generally agreed, but in power amp input stages gridleak bias may be too variable and non-linear, no matter how much I tend to like it.

Of those suggestions only the EL84 is without gridstopper.

Add it.

The scope shows both plates clean as a whistle to 60 MHz at least.

But that is with resistive load on the bench. What about playing music into a speaker?

what do you think of the following for a grid supply?

Bit overdone, but will do no harm....

I tried the E180F with cathode degeneration before converting to gridleak. The gain went up on removing the bypass cap, which I take to be an indicator of Bad Things. It's something I meant to revisit but converted to gridleak. It seems to really like having its cathode hard grounded. Save for the usually SE low frequency stuff the amp's hiss is at the limits of audibility with an ear pressed to the dust cap of a 12" Tannoy Gold.

Something does not sound right here. Did you use the cathode degneration with nthe CCS load or the choke?

Right now the two EL84's share two 60uF and one 40 uf ASC motor runs (for a total of 160 uF)

Sounds fine to me. I did not know how far you took things.

Let's hope! Since the output stage appears to be the main limitation at this point - the front end measures more linear by a factor of 20 - I'll start there.

Yes, I'd think so too.

Sayonara

 

[rdf]

The increased gain with degeneration could have been oscillation above 60 MHz. Wiring in the silver micas last night I noticed g3 was tied directly to plate in converting it from pseudo triode. No stopper resistor. Since it's something I wanted to try anway it was connected to g2 with 820 ohms (value at hand), g2 remains connected to anode with 1000 ohms. This one change dropped the plate voltage from ~110 to ~70 vdc. I recall reading somewhere oscillation can manifest itself as an increase in gain. That would certainly explain both behaviours.

There was just enough time last night for a quick listen. Interesting indeed. The changes appears to have had a substantial effect at the extreme top. Sibilants are very fast, distinctly different between male and females, individuals and location and have a greatly reduced tendency to shift from 's' to 'sh'. Sounds promising.

Thx again.

 

[alexg]

RDF,

Would you mind sharing the schema of your E180F/EL84 circuit?

I want to do an EL84 amp and I have a pair of 6688 (E180F) to try as drivers.

Thanks.

 

[rdf]

Hi Alex. Sorry, that was a transitory design long ago dismantled. The general topology would have been 5K:8 transformer, 300-325 VDC on the EL84 plate at ~32 ma (I prefer high and cold for the reduced high harmonic distortion components), an IXYS CCS on the plate of the trioded E180F driver and an LED cathode load. My guess is a green LED with ~2 ma plate current. That circuit has been rebuilt so many times it's a now a 9002/6CW5!

 

[alexg]

Hi Alex. Sorry, that was a transitory design long ago dismantled. The general topology would have been 5K:8 transformer, 300-325 VDC on the EL84 plate at ~32 ma (I prefer high and cold for the reduced high harmonic distortion components), an IXYS CCS on the plate of the trioded E180F driver and an LED cathode load. My guess is a green LED with ~2 ma plate current. That circuit has been rebuilt so many times it's a now a 9002/6CW5!

Thanks.

I will try a triode strapped 6688 driving an EL84.

I just finished an triode strapped EF86 driving an EL84 in parafeed and I like what I am hearing.

Thanks.

 

[tiefbassuebertr]

SHINDO LABORATORY MODEL 454 MAZERIS BELLEVUS Stereo Pre Amplifier

To prepare this vintage preamp for an external dual mono power supply I need a schematic diagram. Who can call an URL for download ?

about
???????? McIntosh/JBL/audio-technica/Jeff Rowland/Accuphase/?/?¾?????
you will find some pics about this preamp.

Thank you very much.

 

[tiefbassuebertr]

Schematic now I have create

here you will find it as PDF attachement (drawing and measure errors are not to rule out):
http://www.diyaudio.com/forums/tube...how-works-circuit-6067-12bh7.html#post2350495