Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 3 Next   Go Down
  Print  
Author Topic: Collins 30L-1 tube flashover resulting in T-R relay inoperative?  (Read 17839 times)
N6QWP
Member

Posts: 735




Ignore
« on: September 07, 2017, 05:17:41 PM »

After trying out untested tubes in a 30L-1, one of them (a gassy one that ran purplish) gave a loud bang and bright flashover (without excitation) and blew the fuses.  Tried it again and results were the same=now an obvious short. The outcome, after replacing tube and fuses again, was that the relay no longer works and only the exciter rf comes through the amp.

Tried keying the relay line by grounding it directly and still no action on the relay.  

Working with limited knowledge on this, my first amp, and would appreciate any pointers on what might have been affected by the flashover and rendered the relay inoperable.  

Since it was in the receive position when the flashover occurred, might the contacts have been welded shut?  Or some component in the relay circuit failed?  Or the relay itself burned up?  Voltage spike?  Or something else?

Just seeking a direction to go in before I tear into the 30L-1.  Thanx
« Last Edit: September 07, 2017, 05:38:15 PM by N6QWP » Logged
KE7YD
Member

Posts: 100




Ignore
« Reply #1 on: September 07, 2017, 09:43:38 PM »

It probably poped the resistor and diode in the bias supply that's associated with the relay.  Take a look at the two parallel 10 ohm 2 watt carbon comp resistors that act as fuses to the RF cage also.  It may have taken out the cathode resistors too.  The collins collector association has lots of information on their web page.
Logged
W1UO
Member

Posts: 11




Ignore
« Reply #2 on: September 07, 2017, 11:23:11 PM »

The T/R relay coil is in series with the bias supply.  A HV plate to grid short can burn out the relay coil.  The 47 ohm grid resistor can be burnt.  Also the parallel 10 ohm resistors in the power supply.
There is a way to minimize this damage in the future. Check the CCA reflector for the info.
Logged
VK6HP
Member

Posts: 525




Ignore
« Reply #3 on: September 08, 2017, 06:26:31 AM »

You may be lucky and just burned up R28, the 39 ohm resistor in the keying line (the value of which, by the way, establishes the grid bias voltage).  

Since you mention the ARB-704 in another thread, I note that you've almost certainly blown one or both switching transistors in the ARB-704.  If you connected it to a working 30L-1 you'd probably find that it tries to key the amplifier continuously. The original ARB-704 transistors are a bit expensive here in Australia but can be replaced with the much cheaper FJP13009TU type.  When you've fixed your keyer, ignore its optimistic specs, put it in the cupboard and use a more robust variety with the 30L-1.  Either a fast mechanical relay or a high voltage MOSFET type is fine.  I'll note in passing that it's been my experience that external impedance and/or poor bypassing with the 30L-1 keying line can produce a tendency towards the "medium frequency oscillation" excited by power supply ripple, one of two well documented RF instability modes in the linear.

Take care when checking the linear keying directly: if there is continuity through the relay coil and R28, there's -170V at the centre of the phono connector: more than enough to prickle. (I'm guessing you're not seeing that voltage in the present fault condition).

As others have said, check the CCA site for good guidance, especially the "rx for your Collins" section.  In addition to checking the plate and grid components mentioned in previous posts, check also the fusible link from the centre tap of the filament transformer to ground.  In a bad overload, this can open up and, when you've repaired the amplifier, you'll find you're transmitting a substantial hum.

The bad news is that to get to R28 you need to remove the amplifier from the outer case, then remove the back panel.  You'll see the resistor mounted on stand-offs near the Type-N RF output connector.  You can run the amplifier with the back panel removed but note the 3 HV safety interlocks for the other panels: bottom, RF deck and power supply. The HV caps need to have discharged before you remove these panels or you'll be treated to a shower of sparks and potentially more damage.

There's one "must do" modification discussed in the CCA material. While you have the back panel removed,  add a MOV or transorb to the grid supply to limit the grid voltage and subsequent damage caused during events such as the one you experienced.  I used a 200V dc MOV in my 30L-1 but you can read about the merits of the various approaches.

« Last Edit: September 08, 2017, 06:31:39 AM by VK6HP » Logged
N6QWP
Member

Posts: 735




Ignore
« Reply #4 on: September 08, 2017, 06:55:47 AM »

Thanx so much for the informative replies!  I looked at the CCA site, but had some difficulty in finding anything near as detailed there (not familiar with how to navigate there yet).

I found a relay for the 30L-1 on epay last evening and was able to get it for $58.  It was supposedly the last one and I didn't want to let it go by-in case I needed it.  I wouldn't know where else to find one.

Is there a good source for the other components that might need replacement?  MOV?  My local electronics store just closed up and Radio Shuck is also gone.  Moser?  (ARB transistors)?

I greatly appreciate the input given.....glad there are knowledgeable hams willing to share their expertise here in this forum.  It is so helpful to have a place to start when about to try and solve problems in unfamiliar territory.  

As an aside, will the addition of the MOV really protect against further damage should I want to try additional untested tubes in the amp?  Is there any way to check old tubes and not risk something happening again?

Since the 30L-1 has horizontally mounted tubes, I am left to finding and using older US made tubes-rather than the ones now made and available.
« Last Edit: September 08, 2017, 07:17:51 AM by N6QWP » Logged
VK6HP
Member

Posts: 525




Ignore
« Reply #5 on: September 08, 2017, 07:34:19 AM »

Yes, Mouser will have most of the components you need, including the MOV or transorb. I have no first-hand experience of an overload with the MOV in place but it should be a sufficiently low impedance to blow a fuse and protect the relay etc, possibly itself being sacrificial in the process.

I would be very cautious with old 811 tubes.  If you really want classic tubes try and find some 811A NOS types of known provenance.  I would not be swapping random, and likely gassy, very old tubes through my Collins amplifier.  It's true the Chinese 811A tubes are not ideal for horizontal mounting but we're now in risk evaluation territory and you likely risk less damage with the new tubes.  My own 30L-1 has two classic RCA tubes and two Taylor (Chinese) tubes: not ideal theoretically, but working fine.  Keep the drive down to 60W or less, tune quickly and expertly to keep the tubes as cool as possible, and be happy if you see 600W PEP into a dummy load. 

When these tubes get to the end of their life I have some 572Bs which I'll swap into the 30L-1. They include some of the new production run RF Parts Chinese tubes, which have been doing very well in an FL-2100B, also with horizontal mounting.
Logged
VK6HP
Member

Posts: 525




Ignore
« Reply #6 on: September 08, 2017, 08:18:55 AM »

If you look on the CCA web site under the list of 30L-1 documents you’ll find an excellent paper on the care and feeding of the 30L-1.  It details the MOV installation and includes much good general advice.

There is no practical danger of a reasonably sized MOV catching fire in this application.  MOVs can absorb a great deal of energy for a short time and the 30L-1 fusing etc will kick in quickly enough.
Logged
N6QWP
Member

Posts: 735




Ignore
« Reply #7 on: September 17, 2017, 08:08:22 AM »

Wondering about CR17, CR18, CR19 which I have been advised to replace with 1N458's.  Would 1N458A's work as well?  In Mouser catalog, 1N458 is $21.....1N458A is $0.21???  Proper choice???

Also wondering about MOV's to purchase from the Mouser catalog?  Several choices.  Voltages and sizes.

W8JI's recommended 140 volt GDT's in filament circuit?  Or MOV's???

Would appreciate anyone familiar with the Mouser catalog (online) that could differentiate here.   Thanx

« Last Edit: September 17, 2017, 08:30:37 AM by N6QWP » Logged
N6QWP
Member

Posts: 735




Ignore
« Reply #8 on: September 17, 2017, 03:38:00 PM »

After researching 1N458's online, I am stumped by the wild fluctuation in prices....from less than $1 to over $20 at Mouser?Huh!!  Why are theirs so expensive compared to elsewhere?  Are they special and worth it?  Necessary?  Mouser was recommended as a source.

If anyone is familiar with the 30L-1 and replacing these diodes, could I please get some help on choosing the proper ones to acquire? 

I don't want to "cheap out" and get either the wrong parts.....or inferior parts.  I'd just like to know if the $21 ones are what I need--and why.  Thanx
Logged
HAMHOCK75
Member

Posts: 621




Ignore
« Reply #9 on: September 17, 2017, 05:29:55 PM »

Hello, thanks for the Hustler mobile resonators.

To answer your question why the Mouser 1N458 is $21 and the 1N458A is $.20, the 1N458 is a military grade part, the 1N458A is commercial grade.

Mil-grade parts work over a greater temperature range ( -65 to 150 degrees F ) and have an expensive and extensive paper trail so they can trace it back to where, when, how it was made. The $21 part is made by Microsemi in MA. The commercial grade part is made by the Fairchild division of ON Semiconductor. Mil-grade parts are generally not made in large quantities, commercial parts in huge quantities.

Mouser has zero 1N458's in stock but have 2,342 of the 1N458A. Digikey has +80,000 of the 1N458A in stock.
Logged
N6QWP
Member

Posts: 735




Ignore
« Reply #10 on: September 17, 2017, 05:49:03 PM »

Thanx Bill.....my question would then be:  For use in the Collins 30L-1 would the commercial grade be acceptable or desirable?

Is there any real reason to go with the Military version in Ham service?   And not go with the less expensive one?  Are they comparable?

Obviously, I wish to avoid a repeat of damage to the 30L-1.  
« Last Edit: September 17, 2017, 05:54:04 PM by N6QWP » Logged
HAMHOCK75
Member

Posts: 621




Ignore
« Reply #11 on: September 17, 2017, 06:05:51 PM »

The 30L-1 is not a mil grade product so commercial grade parts are what would normally be used and would be perfectly adequate. I wouldn't hesitate to use the 1N458A for $0.20 at all.

Forgot to mention, not all 1N458's are mil grade but the one sold by Mouser for $21 is.
« Last Edit: September 17, 2017, 06:10:36 PM by HAMHOCK75 » Logged
VK6HP
Member

Posts: 525




Ignore
« Reply #12 on: September 17, 2017, 08:04:27 PM »

I would not replace the diodes until you see if it's necessary. It's perfectly possible to have a mild flash-over in a 30L-1 with the only casualty being R28, the bias resistor in the keying line, and a fuse (my experience prior to MOV fitting).  But obviously good to have the parts on hand when you do open it up, just in case.

The MOV is not critical..just choose a 15-20mm diameter part, with a dc rating of 180 or 200V.
Logged
G3RZP
Member

Posts: 1280




Ignore
« Reply #13 on: September 18, 2017, 03:26:10 AM »

With integrated circuits, the difference between 'military' and 'commercial' is generally get enhanced  testing, including over temperature, and on metal can parts, leakage tests. Space grade parts are very different including such things as tests to destruction on samples of wire bonds and tests for anything loose in the package. 25 years ago, a minimum order for space grade parts from GEC-Plessey was  $100k, for which you got 100 parts. Even if you only wanted ten!

Implant quality is worse. A crystal that costs about 5 cents in quantities of 50,000 for cell 'phones is $15 for medical implant use....
Logged
K9AXN
Member

Posts: 442


WWW

Ignore
« Reply #14 on: September 18, 2017, 10:48:00 AM »

Wondering about CR17, CR18, CR19 which I have been advised to replace with 1N458's.  Would 1N458A's work as well?  In Mouser catalog, 1N458 is $21.....1N458A is $0.21???  Proper choice???

Also wondering about MOV's to purchase from the Mouser catalog?  Several choices.  Voltages and sizes.

W8JI's recommended 140 volt GDT's in filament circuit?  Or MOV's???

Would appreciate anyone familiar with the Mouser catalog (online) that could differentiate here.   Thanx



I have a few notes regarding the 30L1 that may be helpful.

There were several changes to that amplifier; a bit scattered.

Go to the Collins site and download the 30L1 8th edition.  2nd page has a list of changes --- but not all.

Download the 30L1 bulletins.  I believe there are maybe 3 changes.

Download 30L1 Instability- causes and cures.

Hidden all over the archives you will find the following and I may be repeating what you find in other documents:

R12 was changed to 2000 ohms.
R28 was changed to 56 ohms These were to lower the idle current.

L3 was changed to from 39uh to 20uh to prevent a medium frequency low level oscillation in 7-7-70.

CR 17, 18, and 19 went through three iterations 1N67, 1N252, then 1N458.  NOTE: Leaky diodes caused the ALC Automatic Load control circuit to drift.

The ALC circuit is sensitive to reverse leakage.  The 1N458 has significantly less leakage than the 1N458A.  If you plan to use the Automat LOAD control function --- use the 1N458.  Many believe the circuit is useless and it is if the wrong diodes are in circuit.  Operating properly, it allows you to take your time to tune the amp up at 300 watts then touch up at full power.  The circuit has to be calibrated when tubes are changed or go soft.   

The 10 meter input tuned circuit was changed in 9-1-67 Adding C75 and C76, 150pf and 100pf respectively for 10 meters.

The 15 Meter input circuit was changed 9-1-67.  C68 and C69 were changed to 220pf and 150pf respectively.

The plate Choke has never been considered for change.  The 30L1 was designed to operate at any frequency from 3.0MHz to 30MHZ.  The plate choke has no resonant points below 30MHz; from lumped constants or velocity.  C16, L13, and C34 provide a low pass filter to protect the power supply caps when using the 80 meter band.

Good luck --- well designed amp Good choice.

Regards Jim
Logged
Pages: [1] 2 3 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!