The 6550 Experience

His guitar slung across his back, his dusty boots is his cadillac.

Moderators: VelvetGeorge, BUG

Post Reply
User avatar
Xplorer
Senior Member
Posts: 2470
Joined: Sun Apr 19, 2009 5:27 pm
Just the numbers in order: 7

Re: The 6550 Experience

Post by Xplorer » Sun Mar 27, 2016 4:31 pm

maybe for the sylvanias 6ca7, but apparently, the 6ca7 existed in the 50' :

http://www.boiaudioworks.com/best-el34-tubes.htm

i couldn't find more infos about these yet.

do you remember cjh's all along the watchtower ? this was the exact sound from AATWT guitar intro, with nothing else involved really.

i hope i'll find some of these old pots because you make me want to try ! i have an old one, RS, maybe 500k, measuring 800k , or i don't remember.

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Sun Mar 27, 2016 9:56 pm

The 6CA7 is a 50s tube. It was American manufacturer's answer to the EL34. As usual, American companies wanted to use existing parts to make the tube to save money, and also to avoid patent issues.
As I have mentioned, the tube is a cross between a power pentode and a beam power tube. It uses a beam plate in place of the suppressor grid (the third grid) as used in the EL34, but it does not use a shadow wound screen grid like the 6L6 and 6550. Electrically it will drop into an EL34 socket. It uses the shorter plate structure of the 6L6 and the 6550 and a fatter envelope for heat dissipation. In fact the plates are the same as a 6550, handy for saving money.
Shadow wound screen grids hide in the shadow of the control grid (grid one), they are concentric wound exactly like grid one so that electrons flying from the cathode to the plate won't see the screen grid and will pass on by without being collected by the screen grid.
The 6CA7 and the El34 have regular screen grids which collect electrons flying on their way to the plate. This means they need to run at lower screen voltages to keep from over-dissipating, and the preferred method was the screen taps on the output transformer, the widely used hi-fi design, and the design of the Marshall Majors and Sunn amps.
At idle the screen and plate are the same potential, but as the plate voltage drops, so does the screen voltage, so the plate potential is always higher than the screen potential, i.e. the electrons are always attracted to the plate.
This makes EL34s and 6CA7s dangerous to use in full-pentode operation like the SLs run them.
Hence the split supplies like I installed on some of the SLs, screens at half the plate voltage---MUCH safer.
And yes the sound is a little different between the EL34 and the 6CA7, but not much.
The beam plate in the 6CA7 helps keep electrons on the plate during large swings of plate voltage.

User avatar
C J H
Senior Member
Posts: 183
Joined: Thu Apr 14, 2005 7:43 am
Location: Sweden

Re: The 6550 Experience

Post by C J H » Sun May 22, 2016 7:11 pm

double post...
Last edited by C J H on Sun May 22, 2016 7:13 pm, edited 1 time in total.
"blah blah, wof wof!"

User avatar
C J H
Senior Member
Posts: 183
Joined: Thu Apr 14, 2005 7:43 am
Location: Sweden

Re: The 6550 Experience

Post by C J H » Sun May 22, 2016 7:13 pm

Xplorer wrote:i'd love to try a set of 6ca7 too, maybe in a black flag, as Jimi perhaps did this on his 67 amp. he got some tones in 67 that sounded very close to what the west coast amp delivers, for example with "all along the watchtower" ( as heard in CJH's clip about the west coast mod. too bad that he removed the clips and doesn't appear here anymore )


Haha, missed me Xplorer?

Nice to see the forum back up and running! Was going to post in the bigger West Coast thread but got fed up with reading all the bullshit the trolls spewed out there.. Damn pity.

Anyway. Like shakti I recently returned my JTM100 mongrel to Weyer-specs. One of the 6550's burnt up a while ago so just running a pair in there at the moment. They seem plenty worn though.. the third one that still lives but is not used has drifted in its matching that it bias 10+ mA below the others. New ones needed and I just ordered a quad of 34's for my hiwatt.. :palm:

So what I did today:

- Put in a pair of 6550's. Biased down to 30 mA and then turned it a nudge hotter to regain a little sustain. Ended at 33 and 34 mA.

- Put an 12at7 back in the PI

- Wired NFB to speaker jacks

- Played around with some cap values: dropped the V1b plate resistor to .0022 uF - added V2a cathode capacitor (330 uF) - dropped PI coupling caps to .022 uF - played around with bright cap values: snipped it and left it that way.

So now I am truly left with a mongrel of Marshall 100w designs from 67-69.. But I think it sounds glorious, the capchanges added some more bite - that I didn't use to care for since I felt it removed the clarity that I wanted. With Dave's recommended mods and the 6550's I find there is clarity in spades and thus lots of room to add a bit more bite and grit.

Then I ran in to some probs however. The heaters measure 2.8 Vac per side.. Doublechecked and I'm running the 240v tap and my walloutlet puts out 236v last time I measured. Time for a new PT?
So to my question. If it is indeed time to fit a new PT I would like to get a JTM100 so I can get the plate voltage up (currently 521 Vdc). Since I also need new tubes: Dave, do you reckon a set of new production of 6CA7s would hold up under those conditions? Or should I just get a new set of 6550s?

Here is a short snippet from tonights little jamsession and a pic from the scene of the crime. Dedicated to Dave, don't mind the ignorance some people seem to eager to boast - keep sharing all the knowledge you're sitting on! Strat - vibe - amp. Shiittty powersupply on the vibe (hence the noise) and a generous touch of verb to simulate the amps natural habitat :jimi:

https://soundcloud.com/pippo1448/dave-w ... -president

Image

How did you all fare with the fuzz-experiments by the way? Is there a West Coast 3 transistor up and running yet?

Jake
Last edited by C J H on Tue May 24, 2016 12:31 pm, edited 1 time in total.
"blah blah, wof wof!"

User avatar
Xplorer
Senior Member
Posts: 2470
Joined: Sun Apr 19, 2009 5:27 pm
Just the numbers in order: 7

Re: The 6550 Experience

Post by Xplorer » Sun May 22, 2016 8:04 pm

How could one doubt anymore that Dave's magic makes your amp Reeeaaaally Hendrixy ?
indeed your clip sounds glorious. makes me think of the isle of whight and some others; glad to see you back ! and the forum too !
about the fuzz experiment we didn't evolve yet, we'll see about that at some point i guess.

User avatar
C J H
Senior Member
Posts: 183
Joined: Thu Apr 14, 2005 7:43 am
Location: Sweden

Re: The 6550 Experience

Post by C J H » Mon May 23, 2016 7:07 am

Xplorer wrote:How could one doubt anymore that Dave's magic makes your amp Reeeaaaally Hendrixy ?
I agree, but still those willing to doubt should be allowed to do so - what I am having a hard time accepting is the way they express their doubt. I mean if they do not agree and have no interest in taking part in the discussions and gear-fiddeling that goes on here, why bother? Its like those parents at kindergarten that want to ban some book from reading hour for all kids just beacuse their kid or they themselves do not like it. Anyway, hopefully things will cool down and we can get discussions, and Dave, back on track.

Back to my prob:
Read up on heatervoltages this morning and found this on the vintageamps forum
From my old Valvo data book:
+/- 7 % for the tranny tolerance,
+/- 10 % for the wall voltage tolearnce
(total +/- 17 % = any voltage between 5.2 and 7.4 volts)
And from the tube datasheet (Tung sol RI 6550):
6.3+/- .63 Vac

So my combined reading (5.6 Vac) is just within range.. :scratch:
But, what effect will lower than optimum heater voltages have in the amp? Sure would be nice to not have to order a new PT. All JTM45/100 (1202-43) types I can find are US-made, so there would be a big chunk of cash on shipping and import fees alone.

And one for Dave: could you maybe elaborate more on how one would go about to build a separate voltage supply for the screens in the amps we are discussing here?

Wheter or not I change the PT, I am planning a complete rebuild of this amp since its intestines are a bit too messy for my taste after the last two years modding. On the add/fix list are:

- sturdier screen grid resistors

- new electrolytics across the power supply board (will stick to 47uFs but switch to F&Ts since the ones in there are a bit too big to fit properly)

- Reverse wiring of main and standby switch (been wired upside down since a rebuild a couple of years back :oops: )

- Wire the speaker jacks ala Hiwatt (.. a bit of dummyprotection)

- The main board I will leave alone except for making the 6550 and class B mods permanent: lower the bias resistors to a 6550 friendlier value (82k or 150k?), replace the piggybacked resistors for the bias supply drop with one of the correct value.

- Maybe rebuild for a separate screen supply at roughly half the plate voltage (?). Alternatively switch to KT88s that are specced to hold up against higher screen voltages.

The sum will hopefully be an amp that retains the sound it has now that in a reliable and proper way operates 6550's in class b.

EDIT: remeasured my walloutlet, 228vac. A 1202-43 would be ideal since many of them seem to have 230vac taps.
Hmm.. is there anyway to calculate what one of those beasts would yield at the plates with a superlead rectifier, or should one play it safe and rebuild the rectifier ala 45/100? :scratch:

Jake
Last edited by C J H on Tue May 24, 2016 12:32 pm, edited 1 time in total.
"blah blah, wof wof!"

Jazz
New Member
Posts: 8
Joined: Sat Mar 12, 2016 2:28 pm
Just the numbers in order: 13492
Location: Eastern Europe

Re: The 6550 Experience

Post by Jazz » Mon May 23, 2016 3:20 pm

CJH

Significantly lower filament voltage than nominal voltage recommended from tube makers (6,3Veff ) means lower electrons emission from cathode ,significantly reduced output power and significantly reduced tube life to,
BTW ,always check filament voltage on tubes sockets terminals but never on power transformers terminals only ,
since there will be always some minor voltage sag caused from wires between PT and tubes , depending from wires diameter thickness .
Blast from the past...

User avatar
C J H
Senior Member
Posts: 183
Joined: Thu Apr 14, 2005 7:43 am
Location: Sweden

Re: The 6550 Experience

Post by C J H » Tue May 24, 2016 12:30 pm

thanks for the clarification Jazz. Is there anyway to get the heater VAC up without changing OT.. I have 220vac and 240vac taps and 228vac in the outlet. A 230vac tap would had been handy...

The PT is an early Metro 1208-03 laydown (before they got two pairs of HV secondaries) and the color-codes don't quite match up. I have soldered to the primary tap that yielded the highest voltage (presumably 220vac?). But if it is the 240vac tap should it not yield a higher than nominal heatervoltage? Seeing that the input voltage is 228vac?

About tubes and the split screen supply. I might try a pair if KT88 instead of a new set of Tngsol 6550's.
Most (all?) modern KT88's are not kinkless tetrodes any but beam pentodes like the 6550's - with a beefed up spec. Tung sol 6550's are specced for max (plate/screen) 600vdc/400vdc.. Not optimal for the way the tubes are operated in a plexi amp, especially not when the plates are over 500vdc and the screens not far behind? The KT88 is 800vdc/600vdc. Being that my platevoltage is currently 521vdc, the KT88 would hold up quite good under this conditions even if I switch PT getting 560vdc on the plates? Given that the low heater voltage is sorted right?

Since I have not decided on tubes/PT I did not order a new set yet - everything else needed for rebuild is ordered though.

Jake
"blah blah, wof wof!"

Jazz
New Member
Posts: 8
Joined: Sat Mar 12, 2016 2:28 pm
Just the numbers in order: 13492
Location: Eastern Europe

Re: The 6550 Experience

Post by Jazz » Tue May 24, 2016 7:16 pm

CJH

- with 238VAC from outlet connected to 220VAC PT tap you should get the highest secondaries voltages , including 6,3VAC tube filaments voltage to , but in case that you can not get 6,3VAC than that PT is sub-optimal or to small for your specific output power tubes conversion .
-Tung-Sol 6550 is not good choice for that sort of operation where G2(screen grids) supply goes over 400VDC , actually that from Tung-Sol recomended G2 voltage is actually maximum allowed G2 voltage , for longest 6550 tube life in AB1 class of operation(read almost B1class) optimum DC voltages are max.600VDC for anode supply voltage and max.300VDC for G2 supply voltage , lowering this supply values to let`s say 500/250VDC output power will be lowered to but tube life will increase , so KT88 is better choice when you have around 521VDC both for anode and G2 supply ,
-BTW check this basic schematic which explain how to get half value from anode voltage for screen grid supply source from HV secondary center tap , in your case if you want to push 2x6550 on the limit (600/300 DC) secondary AC voltage need to be around 2x215VAC with 0,4A max. continual AC current .
Attachments
PS Doubler.gif
PS Doubler.gif (4.32 KiB) Viewed 4110 times
Blast from the past...

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Sun May 29, 2016 7:46 pm

One needn't worry too much about the heater voltage; my experiments showed that over the longest time period that 6.1 volts gave the optimum life. Try to keep the heaters near 6 volts, 5.6 volts works extremely well on pre-amp tubes and has a good effect on noise (which is why you find starved heater current in sensitive voltmeters at the front end, and in some esoteric hi-fi gear)

I built a 6550 amp for Gary Rowles in 1969 using 710 volts on the plates and 350 on the screens. When I checked it out last year, it still had the original four output tubes and was still a powerhouse. It did need some caps though. In the paper called "A 100 watt amplifier with the BO14 super range transformer" published by Chicago Standard Transformer Company, the setup is precisely as described above. The 6550 can operate all the way to 800 plate volts quite comfortably; just remember that the higher the plate voltage the more into class B you need to operate so that plate dissipation is kept in limits. 30 ma idle current at 800 volts is 24 watts of idle dissipation at the plate, which is more class A than class B.

The bridge power supply shown by Jazz is exactly how I got the 50% screen voltage on the Marshalls.
On other units I used a separate supply sometimes derived from a different winding on the PT. You can use a very low current supply for the screens operated this way because they draw very little power.

I'll be around for a while if you have any more questions.

User avatar
C J H
Senior Member
Posts: 183
Joined: Thu Apr 14, 2005 7:43 am
Location: Sweden

Re: The 6550 Experience

Post by C J H » Mon May 30, 2016 6:37 am

Thank you both for chiming in!

And thanks for clarifying my concerns on the heatervoltage.

Could one replace the choke for a big resistor in that schematic (as in Hiwatts)? If so how do I calculate the value for the intended voltage drop?

The way I am set up now is 521 VDC on the plates and my idle current equals around 16w of plate dissipation. I guess one question for dave would be if KT88's would be a better match if I retain the plate/screens setup as in a stock marshall, seeing that they are specced to handle screens up to 600vdc (atleast most of the modern variants I am looking at)? I want to retain the sound and feel the amp has now, but make it more reliable. Thinking of trying KT88's, if it doesn't work out I might rebuild the screensupply.

I don't really grasp how I would go about this in a SL though. Would I wire the 0.5B+ straight to pin6 and the screengrid resistors? Or would it still go to the same point in the circuit (connecting to choke - through choke to OT).

I will try to work out a layout and post - easier for me and probably easier for you to understand my concerns.

Thanks
Jake
"blah blah, wof wof!"

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Mon May 30, 2016 1:53 pm

Even though a tube may have a screen voltage rating of 600 volts, that in no way means that it should be run in pentode mode with 600 volts on the screen. It is the power dissipation of the screen, nominally around 7 watts, that determines what screen voltage you can use. Look at the current draw of the screen as the plate voltage dips below the screen voltage; you'll see the current rise exponentially in the screen circuit to 40, 50 and then 100 ma. Take 50 ma times 500 or 600 volts and you get 25 to 30 watts of dissipation of the screen! No wonder the El34s melted down, 6550s were better, but it is still way too much for them.
Those screen voltage ratings were for "Ultralinear" circuits (the Marshall Major) where the screen voltage falls identically with the plate as the tube turns on all the way because it is powered from a winding tap on the OT. In those circuits you can run the same plate and screen voltage and get away with it, but not in pure pentode mode like the SLs used.

To be perfectly honest, there is no safe way to run the output tubes in an SL with the plate and screens both running at 500 volts. My compromise was screen current limiting resistors of 2.2K, which cause a voltage drop when the tube turns on all the way and limit the power which the screens can dissipate.
A dropping resistor in place of the screen choke can do the same thing, but the preamp stages lose their supply voltage too during high power output.
A permanent voltage divider with resistors can also get a more favorable screen voltage, but why bother when there is a .5 B+ supply already built into the SL?
Well you won't have the power output when you hook them up this way, but if you have a hotter PT, then that will compensate. For absolute longevity, nothing beats the separate supply for the screens. You can run the tap right to the screens, or you can use the choke there and then another cap before the screens (better). The preamp stages can have their own choke and cap off the B+, or a resistor and cap off the main B+.
If you need specific values for any of these methods just let us know exactly what you want to do and what your voltages are.

Jazz
New Member
Posts: 8
Joined: Sat Mar 12, 2016 2:28 pm
Just the numbers in order: 13492
Location: Eastern Europe

Re: The 6550 Experience

Post by Jazz » Mon May 30, 2016 6:03 pm

Jake (CJH)

Here is my proposed schematic for your future SL using two 6550 in pentode mode where I was refer to original Tung Sol datasheet for 100W output using PT HV center tap to get 300VDC for screen grids supply source , anodes currents is 2x50mA or less and 2x5mA for screen grids and for zero input signal ,
actually from my experience control grids optimum negative bias voltage value is in one specific point where screen grids DC current stop to flow , Ig2=0mA ,
S-BY(Stand By)switch disconnect screen grids supply and turn off amp in stand by mode ,
only one series chocke is used just for 300VDC screen grid supply smoothing ,
preamp stages supply are derived from +600VDC line via three series connected resistors 18K+10K+10K ,
OPT Raa impedance need to be around 5K ,
in this configuration excellent sound dynamics can be achieved altogether with very long life of output power tubes ,
of course as Dave already said there is another possibilities for screen grid supply source but I think this one from my schematic is very simple and elegant one for one E.guitar amp,
BTW, long time ago I have implemented this configuration with 4xEL34 and 4XKT90 for one very powerful tube modulator for custom build AM transmitter , with 850V/425V , and sonic result was outstanding ,
only problem I have was with poor quality ceramic octal tube sockets ,sometime they exploded as firecracker , later I switch to quality ones and problem was solved .
Attachments
6550.JPG
(358.5 KiB) Not downloaded yet
Blast from the past...

daveweyer
Senior Member
Posts: 713
Joined: Wed Oct 29, 2014 9:36 pm
Just the numbers in order: 13492

Re: The 6550 Experience

Post by daveweyer » Tue May 31, 2016 12:08 pm

The circuit Jazz posted is nearly identical to the West Coast mod for that configuration. All the values are there and it will work very well. You can use 4 6550s or even El34s for that matter, and they will last and last, even in square wave clipping.
The SL was perfectly designed for this configuration, even through Marshall didn't use it.
I had made the suggestion to Marshall back in 1969 that they switch to this circuit topology, but companies are not prone to listen to suggestions from the field. They just chugged along as they were with all these spectacular amp failures and huge warranty issues. How simple it would have been to change over too, just a cap and a few resistors.
It was Marshall rebuilds which really put a foundation under West Coast Organ and Amp, so I guess I should thank Marshall for not listening..........

shakti
Senior Member
Posts: 2053
Joined: Tue Apr 05, 2005 9:06 am
Just the numbers in order: 7
Location: Ramnes, Norway

Re: The 6550 Experience

Post by shakti » Tue Nov 07, 2017 4:17 pm

Returning to this topic after a long time...I am still looking to get a more dynamic and bolder/cleaner sound from my West Coast mod SL, and am looking into doing something along the lines of Jazz' schematic. Just a few points needed to clarify;

- on Jazz' circuit, it looks like the screens supply is taken from that "halfway" point and run through a choke, correct? It has essentially all the filtering of that first filter stage (50 or 100uF total, whatever you choose to use), then the choke, then another filter cap stage just for the screens. Isn't that somewhat redundant to have all that regulation for the screens?

- the full HV supply goes to the plates, and in the schematic it looks like that is turned on right away, independently of the standby switch?? Isn't that a problem?

- furthermore, it looks like you have to add power dropping resistors to the downstream B+ line to make this work for the preamp, otherwise the anode voltages in the preamp will be very high, correct?


I was thinking about doing it like this to minimize the "invasiveness" of this mod:

- mismatch the primary/secondary on the PT, i.e. Use the 220V tap on a 240V line. This should bring my full B+ up to around 560V which should be a decent voltage for 6550 plates.
- possibly add some voltage drop resistors to the heaters to bring those down to 6V-ish.
- run the screens from the halfway point, i.e. around 280V. These will be served by both the first stage filter (2 x dual 50uF) *and* their own 2 x dual 50uF screen filter (since the amp is already built with that setup). Question; will it offer better regulation to add a small voltage drop/decoupling resistor between these two filter stages? Ideally it "should" have a choke in there, but I don't want to add too many parts.

I have some small problems with ghosting in this amp when running it stock but with 6550s. I know that ghosting can be the result of a choke being run "too hard" outside its current rating, which I suspect is part of my problem when running 6550s with relatively high plate and screens voltages. I am hoping that effectively removing the screens current from the choke's work load (which is what I am proposing above) might reduce my ghosting problem? In fact, only the preamp would be run through the choke if I wire it as I suggest.

Any thoughts on this Dave?
JTM45 RS OT, 1973 18W, JTM45/100, JTM50, JMP50 1986, JMP100 "West Coast", AC15, AC30, BF Super Reverb, Boogie Mk 1, Hiwatt CP103, DR103

Post Reply