From KT90 to KT88

[ManUtd]

Can i just swap them?
My amp uses 4 peaces KT90 and i want to change to KT88.
Do i have to put in one by one adjusting bias seperately?
What`s the difference in bias on these tubes?

Regards,
ManUtd

 

[rknize]

They should swap fine. What amp is this? The KT88 does have slighly lower plate dissipation, so you may want to turn the bias down a bit before putting them in. Do them in pairs or all at once, not one at a time.

 

[ManUtd]

They should swap fine. What amp is this? The KT88 does have slighly lower plate dissipation, so you may want to turn the bias down a bit before putting them in. Do them in pairs or all at once, not one at a time.

Thanks for Your reply.
It is a chinese amp. Push Pull design.
Ming Da MC368-B90

 

[rknize]

Sounds like that amp will bias for EL34s as well.

 

[ManUtd]

Scematic

Sounds like that amp will bias for EL34s as well.

Here is scematic with KT90.
Is it 33k at V1 and V2 ? for Bias adjustment?

What will readings be with KT88 ?

 

[rknize]

Correct.

 

[ManUtd]

Correct.

Thanks,

And those measurement should be similar, no difference in measurement?

 

[richwalters]

220K grid leaks must be a record value with fixed bias around 500V B+. Pray it doesn't get flogged.

 

[ManUtd]

220K grid leaks must be a record value with fixed bias around 500V B+. Pray it doesn't get flogged.

I`ll check with my amp inside.
Found this schematic on net.

 

[rknize]

The amp may have been designed with EL34s in mind. If the coupling caps are 0.47uF per the schematic, you should be able to use 100k gird leaks without losing LF response.

 

[richwalters]

The KT88 was used with the original design; Glancing at the EL156, and KT90 specs, both sim tubes in power; max 100K gridleaks in fixed bias is spec'd and these values would load the LTP phasesplitter outputs and produce more thd than using
220K....sonically one might detect subtle differences with ordinary AB1 music ratings. The real problem with high value grid leaks is they loose control of a hot grid, leading to anode current and Miller C runaway so the low value grid leaks are done for a controlling reason.

But as amp isn't MI, the problem isn't so apparent. The 6550, a common MI tube that gets thrashed, makers recommend 50K grid leaks in fixed bias and using larger tubes the situation gets tougher for the driver. There are commercial designs that use higher grid leak values (which often sees tubes with a lot of life still left, being unnecessarily rejected for poor hot performance) but if one really wants to force the issue out of high power tubes, then the driver circuits have to be addressed to deliver the goods and an LTP isn't the ultimate driving machine. The game changes to that of low impedance and direct drive.

Keep to the spec sheets that apply to the circuit one is using.
richy

 

[ManUtd]

They should swap fine. What amp is this? The KT88 does have slighly lower plate dissipation, so you may want to turn the bias down a bit before putting them in. Do them in pairs or all at once, not one at a time.

How Can they be adjusted all at once? since they have individually pot`s?

 

[ManUtd]

The KT88 was used with the original design; Glancing at the EL156, and KT90 specs, both sim tubes in power; max 100K gridleaks in fixed bias is spec'd and these values would load the LTP phasesplitter outputs and produce more thd than using
220K....sonically one might detect subtle differences with ordinary AB1 music ratings. The real problem with high value grid leaks is they loose control of a hot grid, leading to anode current and Miller C runaway so the low value grid leaks are done for a controlling reason.

But as amp isn't MI, the problem isn't so apparent. The 6550, a common MI tube that gets thrashed, makers recommend 50K grid leaks in fixed bias and using larger tubes the situation gets tougher for the driver. There are commercial designs that use higher grid leak values (which often sees tubes with a lot of life still left, being unnecessarily rejected for poor hot performance) but if one really wants to force the issue out of high power tubes, then the driver circuits have to be addressed to deliver the goods and an LTP isn't the ultimate driving machine. The game changes to that of low impedance and direct drive.

Keep to the spec sheets that apply to the circuit one is using.
richy

So, according to my amps 220k Grid resistors, You mean it`s better changing them for 100K when swapping from KT90 to KT88?

 

[rknize]

They don't have to be perfect from the get-go. My comment was just precautionary. You are replacing a set of (I assume) worn KT90s with a set of new KT88s. They may run hot until you bias them in. Chances are no harm will come to them if you run them hot for a short time, but I was being on the safe side.

If you know which direction your pots turn to lower the bias current, just tweak them 1/4 turn, put the new tubes in, and bias them up. If you don't know, turn them down with your KT90s in there and then swap.

 

[ManUtd]

They don't have to be perfect from the get-go. My comment was just precautionary. You are replacing a set of (I assume) worn KT90s with a set of new KT88s. They may run hot until you bias them in. Chances are no harm will come to them if you run them hot for a short time, but I was being on the safe side.

If you know which direction your pots turn to lower the bias current, just tweak them 1/4 turn, put the new tubes in, and bias them up. If you don't know, turn them down with your KT90s in there and then swap.

Thanks, i misunderstood.But now i see what You did mean.

 

[richwalters]

>Replacing KT88's in lieu of KT90's will immediately run hotter. KT90's require less -grid bias volts for the same anode current. If I'd got this right your amps runs at 500V B+; I recon with KT88's, (with KT90 bias settings) will run at +25% more anode current (each tube). Keep an eye on this, at 500V B+ todays KT88's at 65mA+; some brands will start showing some cherry red on the plates in a dark ambient.
In my earlier days (1965), it wasn't unknown for a GEC KT88 running with cherry red anode to have a life expectancy of 10,000hrs. These days, I'd tend to back off from that figure by half; and using New Issue coke bottle TungSol 6550's have lasted nolonger than 2500 hrs with evidence of grid runaway. So much for gold and carbon plated grids, a pretty miserable ripoff by New Sensor.

richy

 

[rknize]

I just had one of my Tung-Sol 6550 reissues runaway in the Simple SE. It came form a matched quad meant for another amp. I won't be buying them again.

 

[kevinkr]

I just had one of my Tung-Sol 6550 reissues runaway in the Simple SE. It came form a matched quad meant for another amp. I won't be buying them again.

Funny, I was using one as the pass element in the power supply to my phono stage, running at 70mA with 250V across it it lasted barely two years. Needless to say I won't buy any more, in fact I am now running a vintage GE 6550A in that socket.. The reissue died somewhat noisily (lots of gain in the stages it powered) no evidence of electrical damage, but it developed an internal rattle which makes me think that some weld somewhere failed..

 

[richwalters]

Funny, I was using one as the pass element in the power supply to my phono stage, running at 70mA with 250V across it it lasted barely two years.

The funeral pyre is consistent. My tube simply stood up in the orange without reason. Naturally I checked soldering, the obvious and nothing unusual was found. I'd even changed sockets.
If there is an uncorrupted link that works to NewSensor then I will comment.

richy

 

[ManUtd]

Will i be able to measure Bias voltage across cathode resistor 10R/5W ?
This voltage should be around 1V for KT88 meaning 100mA Bias current?

I mean by doing this i avoid from measuring directly on High Voltages (Do`nt like)