Pullen mixer or gated beam mixer?

[K1ZJH]

I'm thinking about building a version of one of Crosby's HBR receivers, with some modifications.
I'm trying to decide if it is worth using a gated beam mixer tube (7360 or similar) or to go with
the 6ES8 version of the Pullen mixer that I've seen used in the 75A5. Or even some 6J6s since
they've been multiplying in my cellar for decades...  I keep vacillating between going whole
hog, by then I keep thinking simple and cheap has some merits too, especially since the parts
are totaling around $75,456 in 1959 dollars.

So, I need someone to talk some sense into me, I am not sure the outlay and work for
the 7360 is going to buy me all that much in dynamic range or noise figure?  Anyone
try both in set and can share their experiences?

Pete  k1zjh

[W1VD]

Hello Pete

Can only speak to the 7360. I've spent the last six weeks restoring a Squires Sanders SS-1R (no rf amplifier, 7360 first and second mixers, i-f filter, i-f amplifiers, etc) to peak operating condition and making receiver dynamic range measurements. I'm in the process of writing a web page on the results.

On 40 meters (only) the receiver is single conversion so that provides a convenient look at the performance of a single 7360 mixer all by itself. The receiver has no rf amplifier so the progression is antenna, front end filter, 7360 mixer, i-f filters, i-f amplifiers, detector and audio.
 
As built by Squires, the LO to the 7360 was about 6V p-p...RCA recommends 10V p-p to the deflection plate. The measurements below are for the stock receiver with 6 V p-p and other levels using an external HP8640B and amplifier as the LO.

Stock (6V p-p):
MDS: -136 dBm
Blocking (20 kHz): 120 dB
Two tone (20 kHz): 89 dB

HP8640B (10V p-p)
MDS: -138 dBm
Blocking (20 kHz): 129 dB
Two tone (20 kHz): 94 dB

HP8640B (14V p-p)
MDS: -138 dBm
Blocking (20 kHz): 132 dB
Two tone (20 kHz): 94 dB

Note the good noise floor right into the mixer. The front end filter is simple and ingenious. It's a combination parallel tuned at the operating frequency and series tuned trap at the image frequency network and exhibits very low loss - thus the good noise floor. Image rejection is a lackluster 60 - 70 dB or so, though. Have a feeling, but haven't measured it, that the relatively unsophisticated i-f filters are responsible for the two-tone dynamic range topping out at 94 dB.   

Hope this info is of some use.

Jay 

   

 

   

[WA1GFZ]

Man that is quiet. I bet you could drop a 20 dB pad ahead of the radio on 40 meters.

[Pete, WA2CWA]

Nice to see the great progress you made with the SS-1R. Some great information. Now you need a SS-1V 

[w3jn]

Pete, I'ma gonna give ya the same answer I gave ya on ARF 

[K1ZJH]

Pete, I'ma gonna give ya the same answer I gave ya on ARF 

One other problem is that Crosby used the second harmonic of the LO on the higher bands
to reduce drift problems in the HBR receivers.  I can see how that might work with a 6BE6 mixer,
but I suspect the Gated Beam mixer wouldn't work in that application..

Pete

[KM1H]

Jay, a couple of things since I spent a fair amount of time with a SS-1R at Sam Harris' place when he was evaluating it when it was new.

-  MDS isnt as important on 40, what is it and the images at 10M?

- Simulate a real world antenna with varying VSWR's up to 3:1 or so.

- Hook it up to a multiband antenna during a lot of activity.

Our experience was that performance dergaded rapidly from a pure 50 Ohm load as that simple filter wasnt up to the task. There were several others that tested those receivers back then with similar results and the writeups may have survived.

The best I could do with a 6GM6 and 7360 in a 75A4 was a 10M NF of a hair over 6dB showing the consequences of the added switching as well as a less than ideal LO level. For all but those rare and extra quiet 10M days it is sufficient and out hears any SS rig I own without using an external 1dB NF preamp.

My current experiences with a 6ES8 Pullen are most encouraging and I expect to finalize and publish the data later this summer. It does require attention to LO level and isolation as well as careful reading of all of Pullen's notes. A 6GM6 RF amp will result in a 10M NF in the 2dB range using an input tuned circuit, one at the output of the 6GM6 and another at the 6ES8 input. Coax connected as Im building modular for ease of experimenting.

Carl
KM1H

[K1ZJH]

Carl

Is this a variation of the Pullen mixer that ran in ER for the 75A4 a while back?

Pete

[WU2D]

I put a pullen mixer in my SuperPro. Now the original 6L7 had two stages of RF ahead of it and I suppose these helped to overcome the noise!

I did not go crazy, I was simply going to use a 6SL7 dual triode but I found aan octal in metal like your 6J6 with a common cathode the 6SC7 and used that.

I have no data other than to say it works great and is not noisy.

Mike WU2D

[W1VD]

Carl

First, I'm not portraying the SS-1R as an example of good or bad receiver design...just reporting on the measurements I've made. 

The MDS on 10 meters (in stock form with starved LOs) is -135 dBm. I'd expect this to improve some with proper LO level as it did on 40 meters.

Image rejection measured somewhat >60 dB on all bands...thus my 'lackluster' comment.

Yes, the manual does indicate that the receiver is expecting something close to 50 ohms.

Until my recent measurements on the 7360 I had no idea what dynamic range performance to expect. A search of the literature and the internet turned up nothing in the way of hard numbers. How about the Pullen? Are there any dynamic range measurements for that circuit? How can one make an informed comparison without this information? Sensitivity and image rejection are only part of the equation.

 

 

 

[Pete, WA2CWA]

Having had a SS-1R for about 16 years, this receiver wants a reasonably resonant antenna for the frequency range you want to receive. Long wires, random length of wire thrown out the window, or antenna impedances with lots of reactance don't work well at all with this receiver. Even multi-band type antennas can be a problem with this receiver depending on the design.

[K1ZJH]

 
Until my recent measurements on the 7360 I had no idea what dynamic range performance to expect. A search of the literature and the internet turned up nothing in the way of hard numbers. How about the Pullen? Are there any dynamic range measurements for that circuit? How can one make an informed comparison without this information? Sensitivity and image rejection are only part of the equation.
 

Jay, that is exactly what started this. I can't find any references for any measurements on
this mixer after weeks of searching. It all seems to be subjective observations so far.

Pete

[KM1H]

Pete, that ER article was copied from a 1967 QST. Here is much of the available info on the Pullen.

http://www.mines.uidaho.edu/~glowbugs/receivers.htm

[WA1GFZ]

looks like a single ended mixer with low Z lo source. Not high performance looking to me in 2008

[KM1H]

Agree completely Jay about measurement data. Since W2VCZ was a very respected DXer, author, and engineer Id have to believe he understood IMD and how to measure.

But as you know its not easy to compare apples and oranges. Im far away from the IF strip and filtering part of my project and Im not quite ready to hack up the 75A4 again.

Perhaps there is a cheap and simple receiver that those interested can purchase and implement the Pullen as well as several different beam deflection tubes. Then share and compare the info.

The various Heath's come to mind as well as the Kenwood R-599 but Im sure there are other choices that wont break the bank.




Mike, interesting about the 6SC7 but I dont know its 10M capability which of course doesnt concern a Super Pro. It has a pretty low transconductance.

Ive used a 7F8 Pullen in a S-40B, with just swapping out sockets, which has been a test bed for several projects and it actually allows 10M signals to be heard! However I havent worked on the LO or RF amp as yet so no real numbers are available but the 7F8 was used quite a bit up thru the FM band.

Carl
KM1H

[WA1GFZ]

Jay,
Correct me if I'm wrong but nobody really measured IMD before the early 70's when solid state receivers appeared.

[KM1H]

Jay,
Correct me if I'm wrong but nobody really measured IMD before the early 70's when solid state receivers appeared.

It was understood and measured at least for the USN as far back as WW2. In the 60's and 70's the high end commercial marine radios had some of the best IMD performance of the time. Racal and Rhode Schwarz were European leaders.

Hams had tried to interest the manufacturers but got mostly lip service. Drake responded somewhat but it was aftermarket development that made the R4C a great contest radio.

It wasnt until the 80's that the contesters mustered enough clout to force the issue. I remember sitting in a lab on several Saturdays in the early 80's when YCCC members brought in their latest rice boxes for IMD and phase noise measurements. There were some pretty savy RF engineers in the club and they constantly came up with improvements. Then shared info with other clubs and finally the ball got rolling.

Another BIG problem was the ARRL who didnt want to run meaningfull tests for many years being afraid of offending the advertisers.

Carl
KM1H

[WA1GFZ]

i worked on a ship board preselector in a rack so they must have been quite interested in IMD but hams were later. on of the masters Rob Sherwood

[W1VD]

Time to set the record straight...

Measurements on amateur receivers began in earnest in the mid 70's with work done by Wes Hayward  (July 75 QST article) in conjunction with Doug DeMaw and myself at ARRL. The new measurement technique spawned 'receiver wars' between the three of us as well as a couple Murphy's Marauders contest ops - Fred Lass, K2TR (CX-7) and Dave Pietrezewski, K1THQ (R4C). Also involved early on with visits to the ARRL were Rob Sherwood (R4C) and Frank Carcia, WA1GFZ (SB-303). 

This link http://www.w1vd.com/receiverspecs.pdf is an interesting original document (circa 1977) of early receiver measurements made at ARRL. Note my old call sign - WA1LNQ - changed to W1VD in late 1977. I made all receiver measurements at ARRL up until my departure in 1980.

The addition of receiver testing in product reviews came about in an orderly fashion once Hayward, DeMaw and myself were sure we had the technique and equipment perfected and the tests were truly representative of 'on the air' receiver performance.       

YCCC (or any other club) had nothing to do with forcing ARRL to print receiver tests in product reviews. Note that the first product review to contain receiver dynamic range tests was the Yaesu FT-901DM in November 1978 QST. This significantly predates anything that YCCC might have been doing in the 80s.

 

   

[WA1GFZ]

A day in heaven way back when. A pair of URM25s and a combiner. I worked on that SB303 for a year to bring the dynamic range from 67 to 86 db. When Jay first tested it for me and showed me the test method I had too much gain in the first IF. Later my boss an old Radar guy Howard Bleam let me use the lab stuff for further testing. 
I think the CX7 is still quite an RX.
Then synthesizers showed up and we all learned about close in phase noise.
Again last night with my new computer running 4 %  horsepower the RA6830 pulled AM signals out of the noise better than the SDR.

[KM1H]

Jay, I dont read anywhere that I claimed the YCCC forced the ARRL to do anything. In fact, once the manufacturers started coming around they went directly to the user for new product suggestions and improvements. Not to say that the end result was always that great

It was a result of many clubs and individuals pooling resources. Im sure you remember some of the discussions at Dayton.

It was also rather recently that the ARRL revised their procedure to include close in IMD testing; I dont remember exactly when.

My own involvement was using Pin diodes in the front end switching of the TS-940's. Initial testing was quite positive but I got a lot of noise from "experts" who claimed the cheap general purpose diodes were fine. A chance phone conversation with Ulrich Rhode, who ordered regularly from RK, turned things around as he had the clout to deal with the manufacturers as well as irrefutable test proof.

I also proved that Pin's in the TX path reduced the broad band residual noise which had been a problem for me when using the 940 to drive VHF/UHF transverters.

Carl
KM1H

[KM1H]

A day in heaven way back when. A pair of URM25s and a combiner. I worked on that SB303 for a year to bring the dynamic range from 67 to 86 db. When Jay first tested it for me and showed me the test method I had too much gain in the first IF. Later my boss an old Radar guy Howard Bleam let me use the lab stuff for further testing. 
I think the CX7 is still quite an RX.
Then synthesizers showed up and we all learned about close in phase noise.
Again last night with my new computer running 4 %  horsepower the RA6830 pulled AM signals out of the noise better than the SDR.

Do you remember what you did to the 303?  There are still zillions of SB and HW rigs in use and Im sure a few would like the info.

As far back as I can remember Sherwood was preaching gain equalization and Ive followed that advice.

Carl
KM1H

[WA1GFZ]

I homebrewed everything between the output of the RF amplifier and IF filter.
The first IF was replaced with a tracking amplifier tuned by VVC diodes. They were tuned by a 10 turn pot geared to the main tuning shaft. Double balanced mixers. Cascaded IF filters and a third set between the IF output and detector. In the end the shielding between stages limited the performance. 86 dB of dynamic range back then wasn't that bad. Heck I think the FT102 is no better. fc

P.S. The new ARRL dynamic range test is bogus to hide rice box close in phase noise problems.  S5 noise floor get real

[W1VD]

Another BIG problem was the ARRL who didnt want to run meaningfull tests for many years being afraid of offending the advertisers.

Carl...just want to dispel any notion that advertisers were exercising influence over advertisement acceptance or product review content...at least through 1980...no idea what policies may have changed, though, after that.

Back to the original question of this thread...

Perhaps what is needed here is a mixer 'test bed' so that various tube mixer configurations can be tested and characterized for noise figure, conversion gain or loss, blocking dynamic range and two tone IMD. Just thinking out loud here...a small chassis, tube sockets, some sort of input and output matching to 50 ohms etc., connectors in and out and for the HV and bias. External adjustable power supplies (like the small HP units) would provide convenient testing under a variety of conditions. The underside of the chassis would have terminal strips to support the required components for each mixer type that could be wired up quickly in a temporary fashion. An HP8640B/amplifier could be the LO. 

I'd be interested in taking on the project as time permits...and probably have most everything in terms of parts and test equipment required. What say...comments, circuits, suggestions?           

 

[KM1H]

Carl...just want to dispel any notion that advertisers were exercising influence over advertisement acceptance or product review content...at least through 1980...no idea what policies may have changed, though, after that.

I have no idea what actually went on especially in the executive offices. Just what Ive heard from a few former members in the T&E area back in the 80's and 90's.

Back to the original question of this thread...

Perhaps what is needed here is a mixer 'test bed' so that various tube mixer configurations can be tested and characterized for noise figure, conversion gain or loss, blocking dynamic range and two tone IMD. Just thinking out loud here...a small chassis, tube sockets, some sort of input and output matching to 50 ohms etc., connectors in and out and for the HV and bias. External adjustable power supplies (like the small HP units) would provide convenient testing under a variety of conditions. The underside of the chassis would have terminal strips to support the required components for each mixer type that could be wired up quickly in a temporary fashion. An HP8640B/amplifier could be the LO. 


We are thinking similar. I have all the necessary lab equipment here including a HP-8970A


I'd be interested in taking on the project as time permits...and probably have most everything in terms of parts and test equipment required. What say...comments, circuits, suggestions? 

As time permits is the key. Ive been poking along on 2 receiver projects for about 6 months now in between work that generates a profit and some AM rig building.

Id suggest we work thru a full set of beam deflection tests with all the available tubes and in all configurations. Im a bit concerned with linearity in some of the less well known tubes.

The same with the Pullen but there is a much larger tube selection plus at least 4 sockets.

Carl
KM1H