Repair job part #2

[Phil Allison]

Phil Allison wrote:

--------------------

One that comes to mind was a MusicMan hybrid guitar amp - the RD50.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf

------------------------------
Question 2:

The same amp has a shameful design problem with the output stage.

Wilful stupidity.

Bet nobody gets it.


** Last chance, I'll post the answer tonight if no-one here gets it.

A predictable event causes mass destruction and defeats simple repair.



..... Phil

 

[Phil Allison]

Phil Allison wrote:
--------------------

One that comes to mind was a MusicMan hybrid guitar amp - the RD50.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf


------------------------------
Question 2:

The same amp has a shameful design problem with the output stage.

Wilful stupidity.

Bet nobody gets it.


** Last chance, I'll post the answer tonight if no-one here gets it.

A predictable event causes mass destruction and defeats simple repair.

---------------------------------------------

** OK - time gentlemen please !

The predictable event is that output tubes wear and fail, often in a short time when used in guitar amps.

Several failure modes are common, including the tube developing an intermittently short from screen to control grid or arcing internally from plate to cathode.

If glass cracks and air gets in, the tube lights up like a neon sign and draws a huge current.

A tube fitted out of key in its socket can short plate to cathode using the internal heater.

Each of the above with destroy Q3 and/or Q4, due to overvoltage or second breakdown. IOW makes it go short C-E.

If the owner fits new tubes after the fuse blows, at least one will be destroyed almost instantly. So it it's off to the repair tech.

Replacement SJE1692 devices were never easy to find and now long obsolete, no recommended sub exists since the SJE was a special.

TIP31Cs are not up to the job and even MJE15030s are a bit low on Vce.

The same disaster does NOT happen with MM tube drive models or other tube amps in general. Cathodes go direct to ground and a shorting tube just blows the HT fuse. Users are safe to try new ones if they like.

MusicMan even issued a service bulletin recommending old models that came for repair be modified by removing the earlier tube drive cct and replacing with the BJT drive cct for better performance.

100& totally, fucking insane.

For the terminally curious, the BJT board looks like this, 4/5 the way down.

http://www.pacair.com/discus/messages/6/35.html

No-one here looked at the inevitable result of tube failure.

Jut like the idiot MM designer.


...... Phil

 

[bitrex]

On 11/9/19 8:25 AM, Phil Allison wrote:

Phil Allison wrote:
--------------------

One that comes to mind was a MusicMan hybrid guitar amp - the RD50.

https://www.thetubestore.com/lib/thetubestore/schematics/MusicMan/Musicman-RD50A-Schematic.pdf


------------------------------
Question 2:

The same amp has a shameful design problem with the output stage.

Wilful stupidity.

Bet nobody gets it.


** Last chance, I'll post the answer tonight if no-one here gets it.

A predictable event causes mass destruction and defeats simple repair.

---------------------------------------------

** OK - time gentlemen please !

The predictable event is that output tubes wear and fail, often in a short time when used in guitar amps.

Several failure modes are common, including the tube developing an intermittently short from screen to control grid or arcing internally from plate to cathode.

If glass cracks and air gets in, the tube lights up like a neon sign and draws a huge current.

A tube fitted out of key in its socket can short plate to cathode using the internal heater.

Each of the above with destroy Q3 and/or Q4, due to overvoltage or second breakdown. IOW makes it go short C-E.

If the owner fits new tubes after the fuse blows, at least one will be destroyed almost instantly. So it it's off to the repair tech.

Replacement SJE1692 devices were never easy to find and now long obsolete, no recommended sub exists since the SJE was a special.

TIP31Cs are not up to the job and even MJE15030s are a bit low on Vce.

The same disaster does NOT happen with MM tube drive models or other tube amps in general. Cathodes go direct to ground and a shorting tube just blows the HT fuse. Users are safe to try new ones if they like.

MusicMan even issued a service bulletin recommending old models that came for repair be modified by removing the earlier tube drive cct and replacing with the BJT drive cct for better performance.

100& totally, fucking insane.

For the terminally curious, the BJT board looks like this, 4/5 the way down.

http://www.pacair.com/discus/messages/6/35.html

No-one here looked at the inevitable result of tube failure.

Jut like the idiot MM designer.


..... Phil

What about the MJ8503?

<https://datasheet.octopart.com/MJ8503-Inchange-Semiconductor-datasheet-15981606.pdf>

It was available circa 1980 and used on the primary side in some early
switchers like the Apple III PSU.

Trouble is I don't know if they ever made it in a TO-220. Have to mod
the heat sinking arrangement for a TO-3.

 

[Phil Allison]

bitrex wrote:

-------------

What about the MJ8503?

https://datasheet.octopart.com/MJ8503-Inchange-Semiconductor-datasheet-15981606.pdf

** A *Chinese* Motorola TO3, high voltage switch with an Hfe of 7.5 ?

I could hardy think of a worse candidate.

99% chance it is a re-labelled fake.


FYI:

The MJE15032 is suitable, but was not an available typo in the 1980s when I needed them. Requires mods to the bias arrangement as its Vbe is higher than the SJE part, causing sever x-over distortion.

https://au.mouser.com/datasheet/2/308/MJE15032-D-95003.pdf



...... Phil

 

[Phil Allison]

Phil Allison wrote:

------------------------

** OK - time gentlemen please !

The predictable event is that output tubes wear and fail, often in a short time when used in guitar amps.

Several failure modes are common, including the tube developing an intermittently short from screen to control grid or arcing internally from plate to cathode.

If glass cracks and air gets in, the tube lights up like a neon sign and draws a huge current.

A tube fitted out of key in its socket can short plate to cathode using the internal heater.

Each of the above with destroy Q3 and/or Q4, due to overvoltage or second breakdown. IOW makes it go short C-E.

If the owner fits new tubes after the fuse blows, at least one will be destroyed almost instantly. So it it's off to the repair tech.

Replacement SJE1692 devices were never easy to find and now long obsolete, no recommended sub exists since the SJE was a special.

TIP31Cs are not up to the job and even MJE15030s are a bit low on Vce.

The same disaster does NOT happen with MM tube drive models or other tube amps in general. Cathodes go direct to ground and a shorting tube just blows the HT fuse. Users are safe to try new ones if they like.

MusicMan even issued a service bulletin recommending old models that came for repair be modified by removing the earlier tube drive cct and replacing with the BJT drive cct for better performance.

100& totally, fucking insane.

For the terminally curious, the BJT board looks like this, 4/5 the way down.

http://www.pacair.com/discus/messages/6/35.html

No-one here looked at the inevitable result of tube failure.

Jut like the idiot MM designer.


--------------------------------------------------------------

** By complete chance, the owners of the MusicMan brand arranged a few years back for a factory in Indonesia to make NEW MusicMan RD50 amplifiers.

On sale, all over the place including Australia for around $1700 or so.

https://www.bettermusic.com.au/musicman-112-rd-50-guitar-amplifier-combo-1x12-50w

** Did no-one tell them the model is a lemon ?

No published schem available on the web, but I seriously doubt any of the issues have been addressed.

It will have Chinese made 6L6s and some whimpy TO220s in the cathodes.

Oh, but now with a neodymium magnet speaker to save unnecessary weight.

Wow.


...... Phil

 

[Phil Allison]

bitrex wrote:

--------------

** A *Chinese* Motorola TO3, high voltage switch with an Hfe of 7.5 ?

I could hardy think of a worse candidate.

99% chance it is a re-labelled fake.


FYI:

The MJE15032 is suitable, but was not an available typo in the 1980s when I needed them. Requires mods to the bias arrangement as its Vbe is higher than the SJE part, causing sever x-over distortion.

https://au.mouser.com/datasheet/2/308/MJE15032-D-95003.pdf




Well what did you use in the 1980s? Sorry, I wasn't there.

** LOL !!

I used selected on test high Vce and Vbe matched MJE15030s.

Used them too for other repairs, like in those horrible Yamaha 2200 power amps.

Vastly better than the cheap Jap 2SD381s Yamaha used.



..... Phil

 

[bitrex]

On 11/9/19 2:38 PM, Phil Allison wrote:

bitrex wrote:

-------------


What about the MJ8503?

https://datasheet.octopart.com/MJ8503-Inchange-Semiconductor-datasheet-15981606.pdf



** A *Chinese* Motorola TO3, high voltage switch with an Hfe of 7.5 ?

I could hardy think of a worse candidate.

99% chance it is a re-labelled fake.


FYI:

The MJE15032 is suitable, but was not an available typo in the 1980s when I needed them. Requires mods to the bias arrangement as its Vbe is higher than the SJE part, causing sever x-over distortion.

https://au.mouser.com/datasheet/2/308/MJE15032-D-95003.pdf



..... Phil

Well what did you use in the 1980s? Sorry, I wasn't there.

My history books seem to indicate high voltage types were actually
available from the real Motorola then. 7.5 at a couple hundred mA would
be a very unlucky beta for the actual part I think

 

[Phil Allison]

tabb...@gmail.com wrote:
--------------------------

HFE of high v trs was down in the pits.
Used lots of... what was it, the once popular LOPTr,
with a gain of about 3. BU something. 208?

** Yes.

Hfe typical was 20 IIRC.

But worse, the makers put a 30 ohm resistor across B-E to push the Vceo spec up by a few hundred volts.

When tried on a standard transistor tester, Hfe reading was near zero.

I came up with a dirt cheap and simple way that got around the problem, test any power BJT for HFe and Vbe in a second or two, unambiguously.

Was published and sold as a kit too.

Many hundreds, I was told.

Know of nothing else like it.


..... Phil

 

On Saturday, 9 November 2019 23:08:40 UTC, bitrex wrote:

Well what did you use in the 1980s? Sorry, I wasn't there.

My history books seem to indicate high voltage types were actually
available from the real Motorola then. 7.5 at a couple hundred mA would
be a very unlucky beta for the actual part I think

HFE of high v trs was down in the pits. Used lots of... what was it, the once popular LOPTr, with a gain of about 3. BU something. 208?


NT

 

[bitrex]

On 11/9/19 10:13 PM, tabbypurr@gmail.com wrote:

On Saturday, 9 November 2019 23:08:40 UTC, bitrex wrote:

Well what did you use in the 1980s? Sorry, I wasn't there.

My history books seem to indicate high voltage types were actually
available from the real Motorola then. 7.5 at a couple hundred mA would
be a very unlucky beta for the actual part I think

HFE of high v trs was down in the pits. Used lots of... what was it, the once popular LOPTr, with a gain of about 3. BU something. 208?


NT

I don't think any of those early switchers would have worked with
current gains _consistently_ that low.

This Agilent switcher from '75 used a 2N6306 but the parts list listed
it as (Selected) so the problem was perhaps the spread. Could be 3,
could be 30, dunno.

<http://www.hparchive.com/Manuals/HP-63005C-63315D-Manual.pdf>

The modern 2N3606 data sheet lists the min HFE as 15 at 3 amps, max of 75.

 

[Phil Allison]

bitrex wrote:

-----------------

I don't think any of those early switchers would have worked with
current gains _consistently_ that low.

** Course they did.

In TV horizontal deflection, you hit the base with current from a small transformer, with about 1 peak amp available.

Hammered the thing on good.

Turn of was the issue, for which some reverse current was needed to make get it fast and so minimise dissipation.

TO3s like the BU208 could run with almost no heatsink.


..... Phil

 

[bitrex]

On 11/9/19 11:45 PM, Phil Allison wrote:

bitrex wrote:

-----------------


I don't think any of those early switchers would have worked with
current gains _consistently_ that low.


** Course they did.

In TV horizontal deflection, you hit the base with current from a small transformer, with about 1 peak amp available.

Hammered the thing on good.

Turn of was the issue, for which some reverse current was needed to make get it fast and so minimise dissipation.

TO3s like the BU208 could run with almost no heatsink.


.... Phil

To return to the original MusicMan schematic for a sec, it seems weird
to drive power tubes "cascode" in the first place. I don't see any
provision for bias adjustments on the 6L6s. they're basically being used
as a high-voltage pass element to keep the HT off the silicon (well,
when the tubes don't fail that is.)

Was the whole point of the exercise to avoid the expense of matching
tubes and bias adjustment on manufacture?

With matched tubes you could fix bias them both with a zener and AC
bypass cap in parallel the cathode, or a dedicated bias supply and
voltage drive them, could probably drive a pentode direct from the op
amp outputs in that case.

 

[Phil Allison]

bitrex wrote:

--------------

To return to the original MusicMan schematic for a sec, it seems weird
to drive power tubes "cascode" in the first place.

** Oh no, there are many advantages.

Plus just one, big black disadvantage.

I don't see any
provision for bias adjustments on the 6L6s.

** They self bias to match the current sources feeding their cathodes.

Using badly miss-matched or poorly performing tubes hardly matters.

they're basically being used
as a high-voltage pass element to keep the HT off the silicon (well,
when the tubes don't fail that is.)

** Current to voltage converters is one way to look at it.

The direct opposite of grid drive.

Was the whole point of the exercise to avoid the expense of matching
tubes and bias adjustment on manufacture?

** Certainly part of it, plus no tube to feed fully filtered DC power to and far better symmetry when clipping then the prior circuit.


> With matched tubes you could fix bias them both with a zener

** Grid bias must track with the AC supply voltage as it is quit sensitive.

or a dedicated bias supply and
voltage drive them, could probably drive a pentode direct from the op
amp outputs in that case.

** It takes about a 35Vrms swing at each grid.

Cathode drive plus split HT supplies does work very well.

Some overvoltage protection for the drive BJTs is needed.

Drive them hard on of the tube goes bonkers.


...... Phil