• 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.

QQE04/5 Curves

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi,

I've found a few of these Philips dual-tetrodes made for UHF amplification.

However, I want to apply them in an audio PP/Class-B construction, driven well into g1-current with a suitable CF/CCS-driver.
I need some info on expected Plate-Load; that's not immediately available from the Philips data-sheet -> electron Tube Data sheets - Q
How to calculate?
I expect to be running the tubes at around 340V, Fixed Bias
Also, each single tube will be strapped into parallel, so in all 4 x QQE04/5 for a stereo PP-amp.
External heat-sink will be applied to the anode-pins.

rgds,

/tri-comp
 

Attachments

  • QQE04-5.jpg
    QQE04-5.jpg
    49.3 KB · Views: 371
Of course there's the raw formula of

Zout = Va/(Pa/Va)

Va = Plate (anode) voltage
Pa = Maximum plate (anode) dissipation

Here it would be something like 340/(20/340) ~ 6K

How close will that be to an optimum plate-load?
Since I'm going to build a P-P amp. are we talking 12K anode-to-anode?

Help much appreciated,

rgds,

/tri-comp
 
There are given two anode curves at the data pages. First with Ug2 = 200 V and the other with Ug2 = 180 V. I would use the latter, draw the load line from 350 V (extend the scale a bit) to Ia = 200 mA. This will give the anode load for a single pentode = 1,75 kohms. For push-pull stage multiply this by 4, which give you total Ra-a = 7 kohms.

I you parallel tubes, the Ra-a will be halved etc. But, by parallelling tubes, the G1-current required will be doubled too. I would use a MOS-fet driver instead of tube in such case.

By fine tuning the Ug2 and bias, you will certainly get the optimum condition.
Since you will most likely apply NFB, the load impedance will be even less critical.
 
Ooh, this just begs to be driven by a power opamp. I have to see if I can find some...

What do you think about driving the QQE's this way?
I did it before with great success driving 2C34/RK34. They gobble up around 10mA of g1 current each, exactly what a paralleled QQE04/5 requires at the max.
The enclosed schematics shows the same CF/CCS that made the 2C34's sing at 12W/8Ohm & 16W/4Ohm. You won't believe the bass they put out! Something that should be experienced, not just written about :)

rgds,

/tri-comp

EDIT:
I thought I'd better enclose a picture of my QQE-Pyramid.
Beats any Egyptian pyramid hands down!:D
 

Attachments

  • CF_CCS_PP.pdf
    15.6 KB · Views: 125
  • QQE04-5_01_.jpg
    QQE04-5_01_.jpg
    75.7 KB · Views: 299
Last edited:
>>BUMP<< to the above question.

Anyone has an opinion or even first hand experience?

I have become a bit worried about running the tubes at Va=375V and Vg2=250V.
The curves stop at Va=250V I suppose for some reason even if the Class-C mode allows for up till Va=400V, see attached data.
Taking Wa into consideration and not exceeding the max, won't that be safe enough driving the tube in AB2?

I hope someone will chime in with an opinion.
If not; I'll just go ahead and build according to the schematics as the power-transformer has already been ordered. If it turns out bad I suppose I could dump some energy in a series-regulator.

Sorry, had to split the QQE-data in two because of file-size limit on forum.

rgds,

/tri-comp
 

Attachments

  • QQE04-5_pt1.pdf
    121.1 KB · Views: 160
  • QQE04-5_pt2.pdf
    164.1 KB · Views: 94
I read a bit, actually a lot (Seely, Romanowitz and others) and dabbled with the curves.
Some of this tube tech stuff is not easily digestable to me from the bare reading.
I'm the kind of person that need to get my hands dirty to understand.
Please give your opinion on if I did or not; understand that is.
Perhaps I'm too optimistic about what the QQE's will endure?

Thanks and regards,

/tri-comp
 

Attachments

  • QQE04-5_(Va-350V)_a.pdf
    212.2 KB · Views: 92
Last edited:
Its like a miniature 8d21.

Apart from your link not working (Here's a link to 8D21 data: http://tubedata.tubes.se/sheets/049/8/8D21.pdf ) I don't quite follow.
They are as far apart as they could possibly be. They only share the fact that they are both vacuum tubes and being tetrodes.

Btw, a picture of the tube is in the first post in this thread.

Weird maybe yes. Low and wide (Well, short and fat! If that's weird I know a few who would fit that description :D ) on a chassis with drop-through transformers should look nice. And since I have a stash of 20 NOS, they should be put to work somehow.

Any reflections on the posted load-line?

rgds,

/tri-comp
 
Last edited:
Its another dual tetrode in the same envelope designed for rf push pull amplification.

The difference is the 8d21 is much larger. Thanks all.




Apart from your link not working (Here's a link to 8D21 data: http://tubedata.tubes.se/sheets/049/8/8D21.pdf ) I don't quite follow.
They are as far apart as they could possibly be. They only share the fact that they are both vacuum tubes and being tetrodes.

Btw, a picture of the tube is in the first post in this thread.

Weird maybe yes. Low and wide (Well, short and fat! If that's weird I know a few who would fit that description :D ) on a chassis with drop-through transformers should look nice. And since I have a stash of 20 NOS, they should be put to work somehow.

Any reflections on the posted load-line?

rgds,

/tri-comp
 

Attachments

  • 71_2006_117_0012_.jpg
    71_2006_117_0012_.jpg
    87.4 KB · Views: 186
I added the Wa plot for both sections in parallel allowing for 20Watt.

The Load-line slightly crosses the Wa in part, but I've seen it before.
I suppose you could call it aggressive.
Will it have any serious impact on tubelife?

rgds,

/tri-comp
 

Attachments

  • QQE04-5_(Va-350V)_b.pdf
    216.1 KB · Views: 82
I made some load-line adjustments because of considering economics.
OPT's aren't cheap in The EU, particularly in Denmark.
I've been buying Edcor's, VVT's and Hammond and they're all expensive when counting-in the shipping-charges.
In Poland there's a company, INDEL that makes Audio-transformers at reasonable prices. They don't produce at wide variety but I suppose demand and supply sets the level.
What comes closest to my requirements, if the load-lines are correct (I would still like an opinion on this), is the TGL 40/001 that goes for around $50,- ex. P&P.
This transformer is made with an EI102/51 Core. It is 4K Ra-a:8R and is specified 40~16000Hz, 40-Watt. Doesn't sound very Hi-Fi but it may still be OK. Not a lot of places to ask opinions about this XFMR.
Delivery from Poland, at least from one company, is UPS around $11 until 30Kg. That's extremely cheap imo.
All in all a couple of TGL 40's should therefore set me back ~$110,- when they land on my door-step. That's below half price of Edcor's.

I plotted a 4K a-a load-line and made calculations, see the attached.
The 5K a-a load-line seemed agressive and this is certainly no less.
In you opinion will this work or is this too high above Pa-max?
Obviously I could reduce Ua to 250V and keep the 4K load-line almost completely below Pa-max lowering the AB2 output to around 28W from 42W.

Opinions?

rgds,

/tri-comp
 

Attachments

  • QQE04-5_(Va-350V)_c.pdf
    216.9 KB · Views: 69
Last edited:
...In Poland there's a company, INDEL that makes Audio-transformers at reasonable prices. They don't produce at wide variety but I suppose demand and supply sets the level.
What comes closest to my requirements, if the load-lines are correct (I would still like an opinion on this), is the TGL 40/001 that goes for around $50,- ex. P&P.
This transformer is made with an EI102/51 Core. It is 4K Ra-a:8R and is specified 40~16000Hz, 40-Watt. Doesn't sound very Hi-Fi but it may still be OK. Not a lot of places to ask opinions about this XFMR

Opinions?

rgds,

/tri-comp

TGL40/001 is huge.
Same EI102/51 core is used at Indel's 200 VA power transformers. I tested TGL40/001 with a pair of PP JJ EL509 with 120 W.
The only problem is very low primary inductance, only 5 H, but this can be some how compensated with NFB.
I did not measure the leakage inductance, but it is low since the high frequency response was beyond 100 kHz and 10 kHz square wave response is perfect.
 
Yes, it's huge, just shy of 3.4Kg.
Picture and core-specifications attached.

Why would anyone reduce spec's for this XFMR the way Indel did?
I could probably get away with choosing a smaller XFMR.
Unfortunately there's none with the Ra-a I need.

/tri-comp
 

Attachments

  • TGL 40-001.jpg
    TGL 40-001.jpg
    42.1 KB · Views: 152
  • 102_5_6.pdf
    86.6 KB · Views: 78
Last edited:
I tried to answer my own question regarding aggressive design.
Feel free to correct as necessary.

Calculating Peak Power Input class AB2 with Ra-a = 4K.

(2/PI) * Ua * Imax) = (0.64 * 350 * .29) = 65Watt

Output calculation; see above .pdf = 42Watt

The loss may then be calculated as (Pin - Pout) = 23Watt

Since the paralleled, combined tubesections are CCS-rated around 20Watt I really don't think it will seriously influence tube-life if I continue along this path.

Agree...?

rgds,

/tri-comp
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.