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 4 5 Next   Go Down
  Print  
Author Topic: 30L-1 RF INPUT CABLE  (Read 27998 times)
N6QWP
Member

Posts: 691




Ignore
« on: July 11, 2016, 10:00:26 AM »

About to hook up newly acquired 30L-1 for the first time.  Wondering about the interconnecting cable from TS-430 rf out to the RCA rf in on the amp and the need for a specific length of coax and if type matters?.  Have read conflicting accounts of 4' to 15' affecting the stability of the amp?  Probably not operating on 15 or 10.  

Any and all experiences welcomed.  73
« Last Edit: July 13, 2016, 07:14:19 PM by N6QWP » Logged
G3RZP
Member

Posts: 1159




Ignore
« Reply #1 on: July 12, 2016, 01:04:13 AM »

Originally, the 30L1 specified an interconnecting cable between transceiver and amplifier of 20.5 feet of RG-58C/U, and an input SWR of the 30L1 of under 2:1. According to the book 'Single Sideband Principles and Circuits', the electrical length of the exciter tank circuit plus the electrical length of cable plus the electrical length of the input tuning circuit should be a multiple of 90 degrees to avoid phase distortion caused by the reactance resulting from a non-unity SWR.

A number of people have found that especially on 10 and 15, the 30L1 can be unstable with some SWRs less than 2:1, and that because of feedback, the length of input coax can be critical in these situations. Note that the 30L1 does not use neutralisation: interestingly, the RCA Transmitting Tube handbook of similar date that show a four 811A amplifier using grounded grid DOES neutralise them, and some amplifiers using two 572B for 1kW input use neutralising. Note that an SWR of 2:1 allows a wide range of R +/- jX e.g 25 + j0, 30 - j20, 80 + j35 and so on.

I believe later versions (mine is 1962) of the 30L1 handbook allowed about a 4 foot interconnecting cable. The 20.5 foot cable may have been specific to the KWM1, 2 and 32S exciters: the handbook is very coy on the subject of using other exciters.

The output load SWR of 2:1 is fairly conservative: the problem is likely to be that some loads on 80 and 40 will be too far out to enable proper loading of the amplifier. It is possible but unlikely that a 3 or 4 to 1 could lead to the loading capacitor flashing over on peaks. However, the use of a tuner is likely to help ensure that the FCC harmonic requirements are met: the 30L1 is specified as -40dB, while the FCC ask for -43 and the international Radio Regulations ask for -50dB. It is certainly a case that where people have a tri-band (10, 15 and 20m) beam with 8dB of gain on 10, the second harmonic of 20m even at 50dB down can be heard over a wide range - 31mW is QRPP, but still gets out well!
Logged
N6QWP
Member

Posts: 691




Ignore
« Reply #2 on: July 12, 2016, 09:06:51 AM »

Thanx Peter-Wondering where I can go to interpolate the year of my 30L-1 from my serial number of 11120?  Could you provide your serial number and a guess?
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #3 on: July 12, 2016, 11:03:16 AM »

Pete,

How does that cable length work out for all bands???

Regards Jim
Logged
KD0REQ
Member

Posts: 2364




Ignore
« Reply #4 on: July 12, 2016, 07:27:18 PM »

W8JI has written about the "super-cathode" 30L1 and its resonances, among similar amps.  Collins changed their mind a few times on the length of the input coax as different complaints came in about the amp taking off on different bands.  get the story off his website. he has a "thing" about floating the grids.  and plenty of scope evidence to back it up.

as for Joe Operator, if you're running pretty much single-band, and the manual's length doesn't cause the amp to take off, have fun on the air.

if you're a tinkerer, read it all, there is modification information there to bring the grid to ground and get rid of the issue.

another reason to be suspicious of miracle engineering.  better known these days as a blank box with a questionmark in a logic diagram or business plan, followed by "profit!"
Logged
G3RZP
Member

Posts: 1159




Ignore
« Reply #5 on: July 13, 2016, 03:31:47 AM »

Quote
How does that cable length work out for all bands???

Including, presumably, non-amateur frequencies. The answer is "You tell me!" I don't know - I merely quote the manual. Presumably, if one knows the electrical length of the 30L1 input circuits and the KWM and 32S transmitter outputs, one could see if it met the criteria in the book by Pappenfus et. al. At 30 MHz, the loss would be 0.4dB: if the SWR was as high as 2;1, there would an additional 0.1dB. The electrical length 9.46 metres, so about a half wave on 10 and a quarter wave on 20.

So that could be a 'magic' length that happened to stabilise it on some loads.

I have no clue what the serial number of mine is - I suspect it may have been removed. It is currently in bits (as it has been for a couple of years) undergoing a major rebuild. A bit of a 'pig in a poke': as obtained, several of the feedthrough capacitors to the PA were broken, the tubes were of mixed origin and some of them were not even oriented correctly for horizontal mounting, so they did not meet the 811 spec. The tube sockets were very loose and are to be replaced - several of the contacts are broken and the 'spring' has gone from them so that they don't grip the tube pins. A new rectifier board and electrolytics have been done but not yet step-start and a power relay - as is apparently fairly common on 30L1s run on 50Hz 240 volts, the main power switch had welded contacts, which I managed to fix, although getting at it is major job in itself. The power transformer would appear to have been rewound as it now has a 240 volts only primary. I'm fitting a HV 'glich' resistor and a new plate feed choke - a 40 microhenry choke is really a bit too small for 80m where it has 890 ohms of reactance and at full output, some 2400 volts or so peak to peak of RF across it. So I'm using a bigger choke with a bypass capacitor and then a series choke with another RF bypass cap to the HV to keep RF out of the electrolytics. Plus neutralising. The fan is not original, being 240 volt only, so the amplifier has been well gone over in the past.

Had I had chance to open it up, I wouldn't have bought it....but it was at an auction, and it cost me about $200.
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #6 on: July 16, 2016, 07:23:46 PM »

W8JI has written about the "super-cathode" 30L1 and its resonances, among similar amps.  Collins changed their mind a few times on the length of the input coax as different complaints came in about the amp taking off on different bands.  get the story off his website. he has a "thing" about floating the grids.  and plenty of scope evidence to back it up.

as for Joe Operator, if you're running pretty much single-band, and the manual's length doesn't cause the amp to take off, have fun on the air.

if you're a tinkerer, read it all, there is modification information there to bring the grid to ground and get rid of the issue.

another reason to be suspicious of miracle engineering.  better known these days as a blank box with a questionmark in a logic diagram or business plan, followed by "profit!"

Scott,

Regarding Super cathode drive.  I believe your speaking about the url   (http://www.w8ji.com/vhf_stability.htm).  You might think about dropping the numbers from the Heathkit SB-220 grid circuit schematic into your calculator to verify the point of high impedance as depicted in that SPICE chart.  The SPICE application is a great design tool if used properly and by a person that does not have an agenda.  The Heathkit is NOT so poor a design as the author states.

Also, his explanation of the workings of Super cathode drive are incorrect, incomplete, and PATENTLY incorrect for the 30S1.   

Kindest regards Jim   
Logged
G3RZP
Member

Posts: 1159




Ignore
« Reply #7 on: July 17, 2016, 01:02:06 AM »

Interestingly Jim, the four times 811A GG amplifier in the RCA tube manual also uses 220pF from each grid to ground. However, it DOES have neutralising, which I suspect makes a big difference. As was found pretty repeatedly in the early days of HF SS amplifiers, you can easily get a design which is fine into 50 ohms, but into an antenna which has a good SWR at the operating frequency, it can still be unstable because of the SWR at some other frequency. There was a very good article on it in Design Electronics magazine in about 1975.

It's predictable if you have the equipment characteristics to do a Nyquist diagram with any angle  and magnitude up to the maximum rated load SWR covering the whole frequency range over which you have any forward gain .
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #8 on: July 17, 2016, 09:43:32 AM »

Pete,

As you say, the book is absolutely correct.  When using tubes with the inter-electrode capacities of the 811 or 572B in the GG configuration, neutralization is an absolute requirement especially if the grid is semi floated with the 220pf caps.

Collins apparently thought the negative feedback from the 5pf plate to grid capacity in the 811 which is used to reduce IMD could also be used to neutralize the tubes.  Problem is it's INDIRECT negative feedback.  Has a similar effect but the cathode is the element that needs the negative feedback.

The interesting thing about neutralizing a GG amp is you can over neutralize without destabilizing it unlike the grid driven amp.

Seems a perfect solution would have been simply to ground the grids and heavily neutralize it.  The neutralizing feed back would likely have reduced the IMD.

Last thought.  With a Q of 10 to 12 in the final system I would expect the j deviations caused by SWR to be reconciled by the plate and loading caps leaving only the resistance to deal with.  If the resistance is beyond the design limits an antenna tuner is the answer.   

Kindest regards Jim
Logged
KH6AQ
Member

Posts: 7718




Ignore
« Reply #9 on: July 17, 2016, 02:24:36 PM »

Has anyone tried placing a 2 dB attenuator between the exciter and the amp?
Logged
KE7YD
Member

Posts: 100




Ignore
« Reply #10 on: July 17, 2016, 10:22:05 PM »

Later editions of the manual dropped the 20.5 ft cable.

See the Collins Collector Assn. web page for various editions of the 30L1 manual.
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #11 on: July 18, 2016, 08:55:08 AM »

W8JI has written about the "super-cathode" 30L1 and its resonances, among similar amps.  Collins changed their mind a few times on the length of the input coax as different complaints came in about the amp taking off on different bands.  get the story off his website. he has a "thing" about floating the grids.  and plenty of scope evidence to back it up.

as for Joe Operator, if you're running pretty much single-band, and the manual's length doesn't cause the amp to take off, have fun on the air.

if you're a tinkerer, read it all, there is modification information there to bring the grid to ground and get rid of the issue.

another reason to be suspicious of miracle engineering.  better known these days as a blank box with a questionmark in a logic diagram or business plan, followed by "profit!"

Collins had an issue with the 30L1 because it was four tubes, which are unstable without proper neutralization. K9AXN keeps "saying" a GG amp can be heavily "over neutralized" without incident, but that isn't true. Phase shift can move all over the place as impedances change, and even if you get impedance correct **inside the band you are operating on** to stabilize the amp, at some frequency out-of-band it often becomes regenerative.

This is why, if we we want stable and non-critical negative feedback we simply should use a series cathode resistance between the input network and tube cathodes.

If anyone takes time to look at the grid divider circuit Bill Orr and Collins claim is negative feedback, they will find the high value grid-cathode capacitive reactance is shunted by the grid-cathode diode. In the 30S1 this is fine. The 30S1 never goes into grid current, so the G-K diode is open circuit. In amplifiers drawing grid current over part of the RF cycle the G-K diode has significantly less impedance than the G-K capacitance. This upsets the divider in a non-linear fashion as the control grid moves in and out of conduction with respect to the cathode.

Over each RF cycle, as soon as the cathode goes negative enough for grid current to appear, the G-K impedance drops. This makes feedback over the RF cycle non-linear. Good luck on neutralizing out a partial cycle distortion! Just when the tube is conducting and you want peak current, the grid becomes non-linear with feedback. It is the worse imaginable time to inject needless non-linearity.

If you read the phase shift description in the Collins manual, and if you take time to calculate phase shift in the networks and cables, you will find it is not remotely like the manual describes. To do what the manual describes, the user would need a variety of cable lengths, some pretty long.

My own feeling, and this is just based on how the amp behaves (it can easily be made to oscillate on ten meters by simply closing the load control and being tuned through resonance while keyed) is that Collins got into trouble with repeatable field performance, and found that an additional length of coax helped some stabilize certain field applications. It's hard to back track out of a bad design that is sold as an improvement, and it was difficult to add neutralization in that compact 30L1 package, so someone invented a long convoluted tall tale about this degree and that degree added to a cable length delay degree causing a certain negative feedback reflection way back from the exciter.

I challenge anyone to provide a logical technical to provide a logical explanation of how one delay line length with the input networks in the amplifier can provide the claimed phase shift, or to show how a capacitive divider using tube capacitance can provide stable linear negative feedback when the grid is going into and out of conduction.

No wild arm-waving, do an analysis.   

73 Tom

 

 
Logged
N6QWP
Member

Posts: 691




Ignore
« Reply #12 on: July 18, 2016, 09:07:13 AM »

So....what is the solution(s)?  Simplest first, please. 
Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #13 on: July 18, 2016, 09:27:04 AM »

Frank,

Ground the grids and over neutralize --- simple as that

Tom,

Your explanation regarding the 30S1 is patently in error and the 30L1 not much better --- please take a moment to think before you say that I just don't understand. 

Kindest regards Jim
« Last Edit: July 18, 2016, 09:41:09 AM by K9AXN » Logged
K9AXN
Member

Posts: 441


WWW

Ignore
« Reply #14 on: July 18, 2016, 09:53:10 AM »

Dave,

I haven't tested an attenuator inserted into the cable path.  Sounds like a logical solution but the real problem has never been told and has nothing to do with negative feedback. 

Ken,

The cable length has changed more than once including a 2.5 foot extension for the KWM1 to make 23.0ft.  Not one Collins person can clearly described the reasoning behind the cable issue.  Your copy of the manual and schematic has deleted the cable length statement.  Compare the previous schematic to the one you have, the differences may reveal the reasoning for the cable length.  Please let us know if there is or isn't a difference.     

One engineering note claimed that the 20.5ft cable corrected the SWR's varying load on the driver causing phase modulation, not instability.  This is a multi band radio configuration.  How can one specific cable length correct for all bands in a multi band radio??  There is an answer and it has nothing to do with IMD but everything to do with their engineering staffs super efforts to achieve overstated IMD numbers expressed by management.

Kindest regards Jim
Logged
Pages: [1] 2 3 4 5 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!