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: Prev 1 [2] 3 Next   Go Down
  Print  
Author Topic: Neutralising  (Read 13046 times)
K5PB
Member

Posts: 27




Ignore
« Reply #15 on: August 01, 2015, 02:15:03 PM »

Re using 837s and modifying 1625s for GG operation, I guess it's nostalgia time of a sort.  Like Carl KM1H, many years ago I too modified Jan 1625s by cutting a small notch in the base, fishing the beam-plate lead out of the cathode pin and running it into the grid pin and resoldering.  Still have several somewhere out in the garage (of course, hams never throw anything away)  Grin  Seems like I ran across a few 807s that could be modified as well - think they might have been made by Tung-Sol - but most 807s made the beam connection inside the glass envelope and were not modifiable.

Late in the 1950s, when for many of us more than a few bucks was still big money, a friend and fellow ham built up an amp for 20 meters using two 837s driving six of them, in parallel and all in GG.  It was stable with no neutralization, possibly because of the single-band design. The chassis had been salvaged from heaven knows how many previous home-brew projects and had numerous large holes.  In fact, we referred to it as the "potato grader" amp.  It was driven by a CE 20A exciter with BC-458 VFO and fed into a home-brew 2 element quad using cane poles for spreaders.

One late evening we had quite a European DX pileup going on 20 CW (that amp put out about 450W with 1200 volts on the plates - and we were careful not to hold the key down too long), when out of the pileup came a strange call: RAEM.  In true DXer fashion, we worked him, not worrying whether he was a Slim.  It was only much later when we discovered that he was Ernst Krenkel, Hero of the Soviet Union, who had been awarded a special call.  Ahh, those were the days.

Best to all,
Stu, K5PB
« Last Edit: August 02, 2015, 01:02:53 AM by K5PB » Logged
KM1H
Member

Posts: 5115




Ignore
« Reply #16 on: August 01, 2015, 03:45:20 PM »

Quote
At a guess it came from the concept of the Wallman cascode back in the early 1940s, with the classic 6AK5 - 6J4 low noise amplifier for radar IFs. GG stages in receivers weren't neutralised, but generally used triodes with multiple grid leads and in the 6J4 case, gold sputtered grids to minimise transfer of cathode material to the grid - it had a very small grid - cathode spacing to get the gm up. The other tube often used was 1/2 a 6J6 - why they didn't use both halves in parallel to get higher gm and lower input resistance, I don't know. So receivers could use non-neutralised grounded grid amplifiers and get low noise which requires no feedback - ergo, then so could transmitters! Without recognising the fact that the tubes were very different in construction.....

True Peter but when livelier tubes were developed even in the early 50's triodes required neutralization. That includes the W2AZL 2M converter I built in the 60's, still have, and bring to some VHF/UHF conventions  to amaze folks when it comes up with a 1.2db NF. I dont use it on the air as a slighty modified UK Microwave Modules transverter does a fine job.

Quote
Even the RCA tube manual of 1962 shows an application circuit for a linear using four 811s that has neutralising.

What RC # is that manual?

Quote
I believe that 813s in grid driven linear service benefited from neutralising on 10: a lot of people neutralised them in Class C on 10 as well, but  at least one commercially UK built ham transmitter used one un-neutralised with no problems. No parasitic suppressor and a plate lead of heavy braid - no nichrome in sight! A Marconi tx with two in push-pull Class C didn't neutralise, either. Neither did the RCA ET4336, which used two in parallel.

The WRL Champion 300A I converted to an 813 for a customer needed neutralization on 10 but that was due to the generally poor layout as built.
I also have the January 1954 QST cover rig which is also neutralized. Both are Class C. The ATC/ART-13 didnt need it but only went to 18 mc and I use it often on 17M AM.
That is the limit of my grid driven 813 experience.

Carl
Logged
G3RZP
Member

Posts: 1228




Ignore
« Reply #17 on: August 02, 2015, 12:18:28 AM »

Quote
Even the RCA tube manual of 1962 shows an application circuit for a linear using four 811s that has neutralising.

What RC # is that manual?

Transmitting Tube manual TT5, 1962.

Note the grid of each tube is individually bypassed to ground - with a 0.01 mFd disc ceramic.

Quote
True Peter but when livelier tubes were developed even in the early 50's triodes required neutralization.

In grounded cathode, yes. The 6J4 was meant for grounded grid and has a gm of 11,000 micromhos. The later 417A was up to 27,000micromhos: that meant in gg, its input resistance was down to 37 ohms, as opposed to 90 ohms for the 6J4. The 417A enr would be  93 ohms while the 6J4 would be 227 ohms. But neither would need neutralising in gg, and in a cascode circuit, would have a nice low input resistance to minimise the plate swing of the neutralised grounded cathode stage. That was done to minimise Miller effect reducing the input resistance of the stage.....very clever.

The 6CW4 was popular at one stage in grounded cathode with neutralisation. Not so much fun for a wide tuning range, though, which was an advantage of grounded grid.

Some transmitters, such the Standard Telephones and Cables 80kW tx of the early 1960's, used a gg triode final, giving about 11dB of gain, no neutralisation and the grid firmly grounded - through a grid ring.
Logged
K5PB
Member

Posts: 27




Ignore
« Reply #18 on: August 02, 2015, 02:01:26 AM »

Peter, I have a copy of RCA Transmitting tube manual TT4, dated 1956, and the four-811A circuit you mention is not present.  Must have been added for the TT5 edition you have.

Just to set the record straight, and contrary to what has been claimed elsewhere, the Collins 30L-1 was not a copy-cat version of Bill Orr's super cathode drive circuit.  It was designed by Gene Senti, W0ROW (now SK) as a desk-top companion to the S-Line and KWM-2.  The grid circuit was designed so that drive requirements more closely matched the output of those units.  When I joined Collins in 1961, my first assignment was to work with the Amateur Equipment Design group, which Gene headed at that time, and to write the manual for the new 30L-1.   I saw the 30L-1 tested thoroughly on an advanced spectrum analyzer, and never saw any IMD problems.  Nor was any "magic length" of cable to the exciter required (that was a means to get a little better IMD performance out of the 30S-1, which was already better than nearly anything else on the market at the time).  I also used a 30L-1 extensively on 10 meters and never saw any sign of instability.  For what it's worth.

Cheers,
Stu, K5PB
Logged
G3RZP
Member

Posts: 1228




Ignore
« Reply #19 on: August 02, 2015, 03:58:44 AM »

The book tells me that the cable supplied is 20.5 feet long for the 30L1, although admittedly it says other lengths can be used. The AC connections for 230 volts in the book are, to say the least, weird for the US and lethal elsewhere - putting the two sections of primary in series across the 230 volt line is OK, but connecting the junction to the equipment case and the neutral is to my mind, plumb crazy. The transformer now attempts to equalise the two line to neutral or ground voltages, and the neutral is often not at ground, even in the US. So even back in 1961, it wasn't a good ploy.

The plate choke is far too small for 80m, too, and I don't see why. It has the best part of an amp of RF in it, and the 'cold' end bypassing is not very much, so there is a non-negligible amount of RF current impressed on the electrolytics.

There were a number of 30L1 used by airlines at Heathrow at one time. They weren't very reliable, which wasn't their fault - the people using them would change frequency and 'forget' to tune up the amplifier. Too technical for them......One other apparently common fault on 230 volts (even when wired properly) was the surge current on switch on gave the power switch a short life. This could conceivably been worse on 50 Hz.  The KW Electronics KW1000C replaced a lot of them. That had two 572Bs, neutralised and 4 preset channels. Even then, some of the operators could manage to omit moving the switch to the right channel. When driven from a KW2000C 4 channel HF transceiver, an interlock could be fitted: eventually, there had to be an idiot lamp attached so that it reminded them to get the channel on the amplifier and the transceiver aligned!

A company I once worked for made police radios; they had a saying that 'nothing can be made cop proof', but the same idea seemed applicable to airline operations people!

It is a fact that the majority of tube amplifiers today are cleaner than the solid state rigs that drive them, and that we have 'advanced' from the stage where higher order products than fifth were negligible to the point where even the 11th order products are still visible on not too advanced a spectrum analyser.

73

Peter G3RZP
 

It's a good few years now since I've been to Cedar Rapids and stayed at Stouffer's. I visited Rockwell-Collins quite a few times, and was invited to give two presentations at the 1986 IEEE Fallcon exposition there, for which Rockwell actually picked up by some devious route the travel tab from Chicago.
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #20 on: August 02, 2015, 04:06:39 AM »

Stu, Collins never should have  should have known better than to float the grids in the 30L1 up from ground like they did in the 30S1. They also should have neutralized the 30L1, instead of coming up with all that nonsense about coax cable lengths and phase shift.  The 30L1 can be made to oscillate on higher bands just by closing the load control. That's why it is overly IMD sensitive to cable length.

The 30L1 is one of the only four tube 811 amps without neutralization, and to make it worse they lifted the shielding grids on low value mica caps. That worked in the 30S1 because it was a GG tetrode with a screen for I/O shielding. It was a terrible idea in the 30L1.

73 Tom
Logged
KM1H
Member

Posts: 5115




Ignore
« Reply #21 on: August 02, 2015, 05:32:59 AM »

Peter, thanks for the TT5 info, guess I need to find one to add to the data bank!

Also thanks for the correction on those old VHF tubes and amps. Its been so long since I looked at the W2AZL converter schematic I forgot what the RF stage was other than a 417A. I still have some NIB 416B and C's and have to figure something to do with them besides selling to a gold scrapper.

Carl
Logged
VE3LYX
Member

Posts: 814




Ignore
« Reply #22 on: August 02, 2015, 05:39:04 AM »

I thought the reverse. Being raised in  the house of a radio design engineer of some experience I learned early on he was no fan of GG design and thought it crude and suitable for amateurs perhaps and would say so at every available oppourtunity. ("Why wouldn't you just build a real amplifier?")That is his problem.
Whenever I build one I always install a stiff wire stub (or stubs)for neutralization similar to what Heathkit did with the DX60B and a 6146B.  It seems to be just enough to keep everything out of trouble and can be adjusted by bending its position. In fact I have them also on my HB twin 811A GG amp. I have over a dozen 1625s . Some are internally connected and some are not. 807s are harder to find. I needed a pair a couple of weeks ago for my WS19 but could only buy  the 5933 tubes which work but are shorter requiring on the set a slight modification of the plate cap retainer. I have wanted to find a couple of scrap ARC5 tx chassis to build a quad four 1625 Amp for my ARC5 station keeping it looking very ARC5. I run mine at 770 volts but series cathode modulate so net voltage is more in line. BTW several of the old ARRL hand books have 807 amps. Have also tried 829B for a parallel pair with some success. Similar to a dual 807.
don
Logged
VR2AX
Member

Posts: 1013




Ignore
« Reply #23 on: August 02, 2015, 05:49:53 AM »

Didn't the Science Museum in Kensington run a 30-L1 back then when most were less equipped. I don't remember their having a big issue ego aside. But at what frequency is neutralization really an issue?
Logged
KG7SWP
Member

Posts: 90




Ignore
« Reply #24 on: August 02, 2015, 06:11:43 AM »

Since this seems like an interested group, I have a pair of NIB 1625s for anyone that needs them. $10 shipped for the pair. I would just like to see them used by someone in need.

Ed
Logged
KM1H
Member

Posts: 5115




Ignore
« Reply #25 on: August 02, 2015, 07:12:39 AM »

Quote
Have also tried 829B for a parallel pair with some success. Similar to a dual 807.

Never used one on SSB but did run the sections in Class C PP on 6M AM and CW in the late 50's and those neutralizing stubs were popular in articles then.

I do use a 60's Hallicrafters HA-6, with a much modified RX section, which uses the later 5894 without neutralization but the grids are swamped with a pair of 3.3K resistors and the stage is very stable. The IMD of the complete TX path is a respectable -36dB after a few changes and the 5th and later fall off rapidly.

Ive also used the cute 815 on 6M SSB in a HB TX converter in the roughly 65-83 years.

Carl
Logged
G3RZP
Member

Posts: 1228




Ignore
« Reply #26 on: August 02, 2015, 09:50:12 AM »

The 5894 is also known as the QQV06-40A in the UK and the QQE06-40 in the rest of Europe and is internally neutralised: adding external neutralisation can make it oscillate. It is the replacement in new designs for the 829B. The same neutralisation business applies to the 6360 aka QQV03-10 and the QQV02-6, of which the US number I cannot immediately remember. There's a little brother, the 6252/QQV03-20 and QQE03-20, which is roughly an 832 replacement.

I think that buy now, all the 832s and 829s have gone soft, as the seals weren't meant to be that brilliant - those made in WW2 were looking at a short life before the aircraft was shot down. It was generally reckoned that a Mosquito had a life of no more than 30 operations in Europe: in India apparently, they weren't shot down in the same numbers and the life was shortened by the glue giving up and the wood shrinking and rotting in the heat and humidity!
Logged
K5PB
Member

Posts: 27




Ignore
« Reply #27 on: August 02, 2015, 01:06:49 PM »

Okay, gentlemen.  The purpose of my comments re the 30L-1 was not to start a controversy, but to offer observations of one who was there at the time and participated in bringing that equipment to market.  Could the design have been improved in some ways?  With the hindsight of decades, no doubt.  We live in an evolutionary world, and that most certainly applies to engineering. 

The suggestion that a particular length of coax between exciter and amplifier could be beneficial came later; it was not in the original manual or the first few revisions. It did indeed originate with the 30S-1.  A mantra in Collins engineering was to continually seek ways to improve performance, even after a unit was well into production.  Collateral engineers were assigned to do just that.  Regarding the inductance of the plate choke, it may very well have been selected to avoid series resonances at higher frequencies, since the 30L-1 as it evolved was intended to be usable on more than just the ham bands.

The quest to make equipment "operator-proof" has gone on since the dawn of the electronic age (with frequent failures).  Today's sophisticated protection circuitry testifies to that, but I doubt if anyone would deny that it has also added complexity and cost to modern amplifiers.  Designers have always faced the impossible task of anticipating and allowing for all the ways in which a product can be used or misused.

The bottom line is that I wonder why it is necessary to continually bash a 60 year-old design and the company that produced it, with exclamation points yet.  Sound arguments have been made and accepted.  Why beat it to death?  I rest my case.

Best to all,
Stu
Logged
G3RZP
Member

Posts: 1228




Ignore
« Reply #28 on: August 02, 2015, 01:35:29 PM »

The big problem with the 30L1 in airline use was that the poor thing was never intended to be used by total incompetents, and one can't bash the design for that!

They were probably managers.........I had 18 managers in 32 years. The good ones lasted the shortest times: the one whose claim to fame was a conviction for gross indecency in a public lavatory with two other men, followed 18 months later by a DUI in a company car, lasted for years! One manager, who didn't last long, in my annual assessment wrote "Peter does not suffer fools gladly, especially when they are in management". It summed matters up very accurately, but I cannot understand why I managed to last for 32 years.....
Logged
W8JI
Member

Posts: 9748


WWW

Ignore
« Reply #29 on: August 07, 2015, 08:14:19 AM »

It's an amplifier, not a person or a company. Technical facts about one design issue do not make an entire company corrupt.

Four 811's in parallel either require neutralization or some very heavy resistor loading on the input to be stable on higher bands. This is just the way it works with the gain and poor shielding inside the tubes.

You can get away with two GG 811 or 572 tubes and no neutralization, three is marginal, and four will be a problem with no neutralization.

The second issue is that small value grid capacitor that migrated through the entire amplifier community. The use of that capacitor was well documented. It was a good idea in the 30S1 because the amp had no grid current, and was a GG tetrode. When that idea migrated into the 30L1 it was a bad idea, no matter who or how it migrated in.

My first hand knowledge of this circuit migrating into other amps came from dealing with Bill Orr, and was through a few companies building amps. Orr was an applications engineer at Eimac. Orr pushed the bad idea of partially floating the grid in GG triode amplifiers of all types, even class AB2 with grid current, as a "super cathode drive". He published various circuits many places, and called manufacturers to push them into using that circuit.

Testing showed the circuit did not work as claimed, and actually made IM performance much worse under certain conditions. It also decreased stability.

Any good logical analysis will show the same.

Why these facts are taken personally puzzles me. It is just how the system works. Anyone who has worked with 811's or 572's realizes they are unstable tubes, and they have considerable output to input feedback capacitance. This is a design issue inside the tube, with the long thin grid wire to a single pin, the long grid, and the exposed direct filament to anode capacitive coupling path inside the tube.

Aggravating that problem through use of a low value grid-to-chassis coupling capacitor and not using neutralization is a bad idea. Grounded grid triodes need to have the grids grounded as reasonably well as possible, and some even need neutralization. The place for NFB is in the cathode lead at the tube, and it should be a broadband stable feedback. \

Orr's push to put what was a good idea in the 30S1 and a very bad idea in the 30L1 universally (that Orr hung various names on over the years) into every GG amplifier everyone manufactured is really the root of the problem, and why that circuit appeared everywhere.

73 Tom
Logged
Pages: Prev 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!