8 Watt SE Plexi Style Tube Amp Made from an Akai Roberts Reel Monoblock

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I have documented the teardown and rebuild of one of these Akai Roberts tube monoblock amps from a reel to reel machine. I have time lapse footage of the teardown, desoldering and cleanup here:

XFMR DIY Projects: Time Lapse Footage of Akai Roberts Tube Monoblock Teardown, Desodler, and Cleanup

Then after the faceplate is painted, labeled, and clearcoated here are some more time lapse videos of the full assembly:

XFMR DIY Projects: Akai Roberts Tube Monoblock to Plexi 8 Guitar Amp Conversion

Here is a picture of another one I built previously. I will have some sound samples soon.
 

Attachments

  • IMG_1359.jpg
    IMG_1359.jpg
    662.6 KB · Views: 401
Here is my hand drawn schematic. If anything is unclear please let me know. By the way, I know folks hate using oddball tubes, but the 6AB4 is just half of a 12AT7. If you want you can just use one section of one of those tubes. You can leave the other section unused or find another function for it. I was going to mess around with a switch that uses both sections in parallel and I can switch one section on or off.
 

Attachments

  • akaiplexi8.jpg
    akaiplexi8.jpg
    194.9 KB · Views: 355
Last edited:
Typically under hifi conditions, with the B+ available a single EL84 is good for 3.5-5w in pentode (my RH84 clip at 3.9 w into an 8 ohm resistor).

Well, this amp certainly isn't hi-fi! Nothing earth shattering in the design. I sort of cobbled together a 5F6-A tone stack with a tweaked pair of differently set up preamp channels and then fed it into a tweaked Akai power section. Looking at the Marshall 5 watter that just came out I like my version a lot better!

The way I came up with the name Plexi8 is, even though it is more like a Bassman tone stack, I like the Plexi name better is all. Same idea, just tweaked. And I have tweaked it back and then made other changes. The 8 came from me working with a friend when building the first one. I told him it was something like 5-6 watts or so and when we plugged it in it was so loud he said that it HAD to be more wattage than that. A Pair of these EL84's has been claimed to have up to 20 watts with fixed bias, so I thought it wouldn't be too much of a stretch to call it an '8'.
 
Typically under hifi conditions, with the B+ available a single EL84 is good for 3.5-5w in pentode.......Most stated wattages for guitar amps are kind of arbitrary to begin with anyway

Not sure what your B+ is using a 250-0-250 transformer and a 6X4 rectifer, or the impedance of your OPT. I can say that 8 watts out of a single EL84 in pentode is possible but I was feeding it about 360 volts of B+ and used a 5K load. Those are RMS watts at 5% distortion at 1KHz.

JJ EL84's will survive on 420 volts of plate voltage as long as you keep the screen around 320 volts. Most others fry quickly at this voltage. I have seen over 30 watts from a pair in P-P with a 6.6K ohm load at 430 volts.
 
Not sure what your B+ is using a 250-0-250 transformer and a 6X4 rectifer, or the impedance of your OPT. I can say that 8 watts out of a single EL84 in pentode is possible but I was feeding it about 360 volts of B+ and used a 5K load. Those are RMS watts at 5% distortion at 1KHz.

My B+ at the 6X4 is around 320 volts and output transformer is a 5k:8ohm I figured the wattage was probably 6.5-7, but I just rounded up. I considered calling it a 6.66 watter, but the 8 just sounds better to me.So, just to be clear, I am not trying to GET it to 8 watts, it was just something I liked and it wasn't TOO far off the mark.

JJ EL84's will survive on 420 volts of plate voltage as long as you keep the screen around 320 volts. Most others fry quickly at this voltage. I have seen over 30 watts from a pair in P-P with a 6.6K ohm load at 430 volts.

Aren't those JJ's really just 7189s labeled as EL84? I know the 7189 is pretty much the same except it can handle more voltage, so that was always my assumption. I have about 100 extra high testing EL84s sitting around my shop from all the Akai Roberts conversions I have done, so I have not had to buy any in a LOOOONG time.
 
Modjeski (Music Reference RM10) gets some 35 watts out of a pair of EL84 running a very high B+ (700v)

That amp has been mentioned here before. When I heard about it, the 700 volt and 12 K ohm OPT stuff was mentioned on the MR web site. It doesn't seem to be there any more. I have been known for turning the knob a bit more to the right than most people, but even I haven't thought of stuffing 700 volts through an EL84. I'm not sure I want to stuff 700 volts through a 9 pin socket since the signal and B+ voltage on the plate could hit 1400 volts, and thats if the amp is treated nice. Bang it into clipping into a loudspeaker and I have seen 3X B+ on some amps.

Aren't those JJ's really just 7189s labeled as EL84?

The guts of the JJ's don't exactly resemble the guts of anything I have labled 6BQ5, EL84, 7189, or 7189A. They have fatter plates with 8 sides, but they have sharp pins like the Russian stuff and copper colored grid rods, with a normal looking getter (not the Russian flying saucer).

There are lots of "stuff" out there (especially on Ebay) sold as 7189 and 7189A that blows up at normal EL84 voltages. An EL84 was rated 300 volts max plate and screen 12 watts plate dissipation. The 6BQ5 has the same ratings. The 7189 has different ratings depending on whose data you read. RCA and TungSol show a 400 volt plate with a 300 volt screen and a 12 watt plate dissipation. Sylvania shows a 415 volt plate and a 330 volt screen with a 13 watt plate dissipation. The 7189A has a 440 volt plate rating with a 400 volt screen rating and a 13.2 watt plate dissipation.

It is possible that vendors like Sylvania just made one tube and stuck all 3 labels on it. I have some old Sylvania 6BQ5's that also work quite well with 430 volts on the plate, but none of the tubes including the JJ EL84 will eat 400 volts on the screen. 350 volts gets the glow of death.
 
Classic story ,on the long run Max 3.5W.

Well, if the Beast makes amps out of old tape recorders, and likes the result, then who cares how much power it makes exactly.....especially when he admits he hasn't measured it. Remember this is a guitar amp.

I once had a Kustom 200 solid state amp. Measuring it into a 4 ohm load revealed 88 watts. Actually it made close to 100 on the initial hit, then the caps ran out of storage, the B+ dropped and the power went down. Watch someone play an old Kustom loud, the pilot light blinks with the music.

If you can only squeeze 3.5 watts out of an EL84, you aren't trying hard enough. THe RCA 6BQ5 and Philips EL84 data sheets show 6 watts in pentode with a 5 K load and only 250 volts on the plate. THe Beast uses 320. 8 watts could be a real number. Probably a bit more when overdriven into a full metal racket.
 
I'm not against the 3.5W at all, here's a Grundig TK46 power amp section with one ECC83 and a ELL80. This one is 1.25W/Ch. I have a Roberts 997 amp pair (same as the Akai M7) ,so after exchanged most of the caps (film and cathode by pass) and the rectifier 6X4 ,it measured 2W into 8 Ohm (4 Ohm less). Just in the heydays of the EL84, in the mid and late 50's most European radios used one EL84 and a EZ80, which is beefier then the 6x4 ,they had put out about 3.5W. In this amp the B+= 291V, Ug2 = 261V ,Uk= 6.7V and Rk = 150 Ohm.
 

Attachments

  • Grundig TK46 amp.jpg
    Grundig TK46 amp.jpg
    180.4 KB · Views: 222
  • Roberts 997amp.jpg
    Roberts 997amp.jpg
    94.5 KB · Views: 217
  • Roberts 997 amp inside.jpg
    Roberts 997 amp inside.jpg
    115.9 KB · Views: 212
After some extensive testing I am getting great tone, but I am also noticing a strange oscillation on some settings. On only channel 2, with gain cranked over 8, the presence over 6 and the treble over 5 I start to get some high frequency oscillation. I tried putting a 47p cap from the plate of V3 to ground and it cut some of it (probably mostly ultrasonic at that cap value), but didn't eliminate it. I was going to try and put a larger value like 270p in parallel with the plate resistor of V3 instead. I was told this a better solution. Any ideas?
 
I tried out the 270p cap in parallel with the 100k plate resistor on V3 and it knocked out most of the oscillation, but not all. Ultimately, I decided to lose the presence control entirely and replace it with a master volume ahead of the 6AB4. SO, now V3 has a fixed feedback loop attached to its cathode and a master volume control of 500k ahead of its grid. While I was originally opposed to the master volume as a tone sucker, I really like the tonal shaping ability that the amp has now. The master volume gives me a wider range of sounds than the presence ever did.

My next unit will have a few more experimental tweaks as I try out swapping the 6AB4 for a 12AT7. One triode the 12AT7 will be set up exactly as the 6AB4 is and the other will be connected to a switch that adds or removes it in parallel with the first triode. Any thoughts?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.