[Top] [All Lists]

[Amps] MKS brand 8877 test

Subject: [Amps] MKS brand 8877 test
From: John Lyles <>
Date: Mon, 30 Jan 2023 10:22:41 -0700
List-post: <>
There was a production issue related to Eimac and the two MRI equipment makers who used the 8877. I attached some information here that I got from another friend who worked in the tube biz then:

ETO was at the forefront of the new MRI diagnostic equipment development and ETO chose to use a pair of 8877 tubes in parallel in their rf amplifier. Their customer was GE Medical. GE Med was having difficulties with the 8877 in their MRI amplifiers: after about 10 to 20 minutes of operation 8877 tubes would short internally between grid and cathode. The staff at Eimac in Salt Lake City had not been able to pinpoint a cause and tubes were being returned as meeting all electrical tests but the would again fail when in MRI systems. It turned out that a small part had been changed in the 8877, a thin cathode heat dam, it was made of nickel in the original design but had been changed to Kovar, a less costly alloy with a low coefficient of expansion but containing iron as one of its constituent metals so rf energy would cause heating and it would expand and touch the grid cap, shorting the grid to the cathode. 8877's returned for warranty tested good at the factory where they passed current division tests (not RF power). As a result many bad tubes were being sent back to customers without the defect being identified. Economic losses at ETO were considerable and ultimately a competitor Erbtec was awarded the rf amplifier contract by GE Medical using a new tube that Eimac developed, a bigger triode. That really send ETO a bad message.

MKS bought one of those companies, sorry I don't remember more. I assume they never used Eimac 8877s after that event. I have a friend who retired from MKS, Colorado Springs. They bought one of the abovementioned MRI machine makers. Eventually had equipment made in China so it was likely a Chinese source triode too. The MRI companies had stringent specs on the tubes when they ran them in parallel. This was something that Eimac in Salt Lake was having trouble making yield with a lot of rejects. Who knows where those tubes ended up too.


John Lyles


Message: 1
Date: Sun, 29 Jan 2023 09:32:01 -0800
From: Jim Barber <>
To: Amps group <>
Subject: [Amps] MKS brand 8877 test
Content-Type: text/plain; charset="UTF-8"

I recently bought a ?mks? branded medical pull. This is my first non-Eimac
8877, so I?m interested in how it performs.

For the sake of the curious, here?s what I?ve found so far:

I gettered it overnight, then put it on my homebrew hipot tester. No
leakage at 8KV, which is unfortunately as high as the box will go.

I installed it in my restoration-project 8877 deck, said a fervent prayer,
then turned it on. No bang, which is always good. The second thing I
noticed is that the heater draws a little more current than my reference
Eimac - It drew ny filament supply (at the tube) down to ~4.8V. After a few
minutes I adjusted the onboard Variac and brought it up to 5V. I checked a
few times after drawing anode current and I haven?t had to change it.

I happened to hit the grid at a low-load time, and the unloaded B+ read
4060V with a Fluke HV probe.

Here?s where it starts to get interesting, for me at least. I keyed the
deck, which applied 12V operating bias to the tube. The ZSAC (idling
current) jumped up to just about exactly 200mA. My reference Eimac draws
150mA under the same conditions.

I went on to do some relatively low power tests, then had to switch tasks.
I didn?t have time to do any structured testing, but I will say that so far
the tube shows noticeably higher gain than the Eimac. How much higher will
be a question for a future session.

The Eimac is, BTW a true NOS specimen and follows the published curves
pretty well. I don?t remember the date code offhand, but can provide it if
anyone is interested.

I found this intriguing enough to order a second tube from the same
supplier. Is the first one a fluke? We shall see.

Amps mailing list

<Prev in Thread] Current Thread [Next in Thread>