• E3443.png

    WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.​

Jadis Defy 7 Schematic

Forgive the "rave" BUT one of my buttons was just pressed.

The Jadis Defy7 has a few problems from a "back to basics" sort of fellow like me. Its just NOT a good design.

The Grid 1 resistors (474K) are too big for 6550 in fixed bias. Any grid current will cause a tube to "run away".
That is:
Grid current flows, causes the coupling capacitors to charge up, that reduces the negative bias at the grid, more anode current flows and therefore more grid current flows and the whole thing builds up until you get smoke (but hopefully not flames).

The obvious fix for this is lower value grid 1 resistors BUT since those resistors are the effective load on the driver section we have another problem. The driver section is just too wimpy to drive the load if we reduce those resistors.

From an AC signal drive point of view things MAY not be quite as bad as first glance would suggest. You would need to confirm this for the Defy7 BUT I can confirm that in the Jadis JA80 monoblocks at least the cathode feedback is POSITIVE feedback and it relies on the NEGATIVE feedback of the Ultralinear connection to keep it stable. This makes the output tubes look like a higher impedance AC load.

It pretty much means that you have to use specially selected output tube sets from Jadis. They are selected not just for matched operating points BUT for low grid current as well.

When I did a restoration on a pair of JA80's this got right "up my wick" and I was determined to find output tubes myself. In the JA80 the problem is slightly worse in that each of the push pull sides is cathode biased and matching of tubes was even more critical. On the other hand they run just 2 tubes (KT88) each side not 3 as for the Defy7. I ended up purchasing 4 matched quads of JJ KT88 to be able to find 2 sufficiently matched quads with low grid current to get the amps running reliably-ish. ('ish meaning that I returned them to the owner running well, but with all my fingers and toes crossed that they would stay that way beyond tomorrow or the day after).

My advise to the guy I did this restoration for was to sell the amps fast while they were working because sooner or later they were going to "blow up" again. The shift to fixed bias in the Defy7 is at least a step in the right direction - otherwise all the warts from the JA80's have just been copied into the Defy7.

If you get the idea that I'm not to impressed with these very expensive examples of so called Hi-End Audio Amplifiers, then you are reading me exactly right. I have yet to see a Jadis Amp I would be comfortable running long term in my system.

Others, if you have an opportunity to purchase a 2nd hand pair of Jadis Power Amps at what seems to be a bargain price, my advice is don't.



Paid Member
gingertube said:
Forgive the "rave" BUT one of my buttons was just pressed.

The Jadis Defy7 has a few problems from a "back to basics" sort of fellow like me. Its just NOT a good design.

<big snip>

Others, if you have an opportunity to purchase a 2nd hand pair of Jadis Power Amps at what seems to be a bargain price, my advice is don't.


I've got to agree 100% with Ian based on identical experience. I've worked on a number of Defy 7 amplifiers and some JA-80s as well. I believe the last time I agreed to work on one was about 10yrs ago. They are a service tech's nightmare for all of the reasons enumerated above, and in addition the Defy-7 is particularly annoying because of the level of disassembly required to replace myriad fried components when the amplifier experiences the inevitable melt down.

In addition to what I will charitably label as at best an "eccentric" design, many of the passive components are underrated for power dissipation, voltage or both.

I will never work on another one of these amplifiers. I desperately want to say something about a group of trained monkeys and that design, but I'll refrain. :devilr:

Edit: The sad thing is the iron is great, the chassis nice, and they even sound pretty good all things considered when working properly. Just don't do it!!!
I happened across this thread, as I was repairing a pair of JA-80s this week in my shop.
They came in with burned PC boards around the cathode resistors, which had been replaced at some earlier date by an unknown tech. Customer bought them in that condition.
Complaint was about a burning smell (gee, I wonder why? <g>) and tubes overheating.
My first order of business was to clean up the carbon mess and seal it so it doesn't produce odors. I changed out the resistors for four 10W 2K ohm resistors on each bank of resistors.
Then I tackled the overheating problem. First, I thought it was gassy 6550s, but out of a batch of a dozen or so, quite a few tubes were exhibiting this problem. So I started measuring grid bias resistors. This amp uses 1M ohm for the bias! Way too high for the 6550. I found that 250K hits the sweet spot, where it prevents run up of grid current, without increasing the distortion of the amplifier at full power output. The amps have been running flawlessly since this modification.
The only other issue I ran into is the RIFA capacitors that they have around the bridge rectifier. One of them exploded after 70 minutes of play, releasing an incredible amount of smoke into the room. Took all night to ventilate the mess. Once things cooled down, I found that these capacitors cracked from age and must have absorbed moisture, thus reducing their already marginal 500V rating. Prior to this, the amp had a low level 60Hz buzz in the output. I removed these troublesome devices from both amplifiers. Now they are quiet--can't hear any noise even with my ear pressed up against the speaker grill.
In short, the JA-80s are ticking time bombs, with numerous things that can fail catastrophically.
Oh, and there's only two tubes in the low level stages on the schematic, but three tubes on the chassis! The third one is just there for show. Only the filament wiring is hooked up--the rest of the tube is not connected! LOL.


Paid Member
Ouch, I have some very vague recollection of working on both two driver and three driver tube JA-80s and had forgotten this little detail.

It has been so long since I last worked on anything commercial the bad memories are fading, in fact I remember very little about those amps - for some reason the Defy 7 has stuck with me, perhaps because it was such a pain to get apart to repair.. I'm done with that phase of my existence.

I remember living in abject fear every time I repaired a pair of these or a Defy 7, lots of work involved and the anticipation that there would be an angry phone call a week or two later when the amplifier self destructed. I ran into a fellow who I did a repair for (major rebuild) some 4yrs later, and to my considerable shock they were still running fine. I had given them a few months at most before the trouble started.

I wanted to modify every Jadis amp I ever worked on, but never really got the green light due to concerns over resale value. I did however change the grid resistors in most of those amplifiers for a somewhat lower value.
The Jadis amps CAN be made reliable, with some design changes. And that should be a resale value booster, because no one wants to buy a ticking time bomb!

The JA-80s have cathode fuses, which cut things off if something goes haywire with the tube conduction. I upgraded to bigger cathode resistors, because the original ones CAUGHT FIRE because they were so inadequate that they acted as the fuse and did considerable damage to the PCB, which was one of the messes that I cleaned up.

The output iron is decent for a modern manufacture transformer, but compared to the really good transformers of the 1950s, such as McIntosh' Unity Wound transformers, they're excessively heavy for only moderate power bandwidth. Full power bandwidth on the Jadis cuts off below 30Hz. Whereas, a McIntosh 240 I have in the shop at the same time, cuts off below 16Hz at full power, while weighing a LOT less. The Jadis look decent today because there are some really awful output transformers being put into VT amplifiers today, like the Carver tube amps. Beware of output transformers weighing in at only 17 oz. The Jadis claims are at least accurate, while the Carver claims are grossly inflated.
bad idea to "upgrade" to bigger resistors - the resistors are sized to prevent catastrophic failures.

You can not compare a McIntosh output stage to a standard output stage.
The Jadis output xfmrs perform well, at least in the Defy 7 pair that I had on my bench.

The only way to make the Jadis circuits work, is as earlier noted to revise the circuit considerably. Then, of course it's not really a Jadis any longer, and the "resale" value may go away, but if you want to listen to them and not have them flame out, that's the way to go.

Imo, which can be discounted, especially on national holidays, which today is one. :D
Some Fixes for Jadis JA-80

I have to disagree on the resistor issue, at least as it applies to the JA-80. The original resistors were underrated, and, instead of the series fuses blowing, the resistors started a fire in the amplifier. How much more catastrophic a failure can you have than a FIRE? This fire destroyed a good portion of the PCB assembly, which I spent time rebuilding.

By upgrading the resistors, the heat dissipation is spread over a larger area, so the point heat temperatures are lower. In series with these resistors are FUSES. The fuses are there to do the job of preventing catastrophic failure.

True, we can't compare the McIntosh output stage to the Jadis. It's in a class by itself, much more sophisticated and has a much better output transformer with a wider bandwidth, much better bass extension (by one full octave) and lower distortion at full power, by an order of magnitude.

WRT the resale value of the Jadis, the heart of the design is in the ultralinear transformer coupling to the amplifier, and, if anything, rational buyers would value a Jadis that has been improved so that it is no longer a ticking time bomb. I do realize that the market for this amplifier falls dangerously into the voodoo market of 'audiophools', but aside from that, I work for customers who have rational needs, and this particular customer wants the amp fixed in such a way that it won't have these catastrophic failures anymore. Based on my using the amplifier for the week after these modifications, it seems that I've accomplished this goal, without adversely affecting the measurable, nor the sonic performance.
Last edited:
After reading the above posts, it only confirmed my belief that Jadis is more of a marketing company - its amps were designed by non-engineers or very bad ones. More here.

Just read that linked page, Patrick Turner has a great style and a great educator.

I was wondering about the 6mm marine ply PCBs. Is flammability an issue? I love the look of them!


New member
2020-03-16 7:54 pm
I need your help to see if it is serious or not.
I was adjusting for the second time the voltage, and putting 5.05v in all the four rows, and suddenly in the resistor I was measuring, a little explosion and a little smoke appeared, and the voltmeter show a value equal to zero in all rows, even the amplifier switches on the light and all tubes are light on.

What do you think the damage is?

Cathode sense resistors?
For measuring the bias?
5V and 100 Ohms is 50mA. Good.
5V and 10 Ohms is 1/2 Amp. Not possible.
What value resistor does your amp have?

But, . . . check the 6550 g1 grid resistors.
If g1 is not returned through 50k Ohms to ground (or in the case of your amplifier, back to the adjustable fixed bias), you will get 6550 tubes doing thermal run-away. Then, Bang! Burn!
The rule is no more than 50K Ohms.

Post # 3 mentions the incorrect 474k g1 resistors.
Post # 5 mentions using 250K g1 resistors (depending on your tubes, your mileage may vary, your 'engine' may burn out).

You might try some very robust (very good) 6550 tubes, and 100k Ohm g1 resistors.

The problem is . . . the driver tube circuit was designed to drive 6550 474k Ohm g1 resistors.
Without a schematic of what you actually have, it is not possible to guess how low of a g1 resistor value the driver can work properly with.
Last edited:


New member
2020-03-16 7:54 pm
Now in any 4 rows, the voltage measures zero, and no sound in the speakers, but all the tubes are switch on, and If I put my ears in the speakers, I can hear the music but very low.

I am Portuguese sorry my English.

But you think it is an expensive fix? It damaged severely the amplifier, or it can be fixed easily?