Bell 2200C Restore--Cathode Feedback

Dave451

AK Subscriber
Subscriber
I'm working on a pair of neat old Bell 2200C amps for a fellow AKer, one of which is being a bit more of a challenge than the other. I've become acquainted with some of the differences between the 2200B model (for which a schematic is available) and the "C" model (for which no schematic is apparently available). The power supply string is quite different and I plan to post the schematic here shortly, showing the differences I've found so far between the C and B models.

However, the most pressing problem is that one side of the output transformer primary is open in one of the units, so we're only getting plate voltage to one of the 6L6GB output tubes (what is the sound of one hand clapping? Well, it is putting out sound!). Output transformer replacement is complicated by the fact that the 2200 amp OPTs have a tertiary winding for cathode feedback (see attached schematic--in this case, the circuitry is the same for the "B" and "C" models).

So, how can I find a replacement OPT? Failing that, is there a work around where I could use a more conventional 6L6 family replacement transformer (Say Hammond or Edcor) in the 20-25W RMS output range and re-work the cathode circuit for the output tubes? The amp also uses a more global NFB circuit from the output side, so that could be increased perhaps.

Any thoughts or suggestions appreciated!
Dave
 

Attachments

  • Bell2200_Bsch.jpg
    Bell2200_Bsch.jpg
    94.1 KB · Views: 92
seems like grounding the cathode, maybe through a smallish value resistor and adjusting global feedback would be the way to do it. Short of a custom transformer wind or an original re-wind I don't know you've got any other choice.

Speaking of which, have you maybe looked into having it rewound? Probably not more expensive to have one rewound vs buying two matching transformers to keep the amps the same. Heyboer and there is another that I can't think of right now are the usual suggestions.
 
*** Following *** I have both the "B" and "C" units sitting on a shelf waiting for restoration.
 
You can get the bad OPT disassembled and rewound. Not sure of cost. I think Heyboer has done this before and there are others. I've never asked into this with a winder. I think sometimes Edcor will make a special on request. Might talk to them.
John
 
I have notes in to Heyboer and Hammond. Edcor responded that they don't do cathode feedback transformers and don't know who does.

I'm studying the cathode circuit of the C model. One difference with the B is that the C uses a 6SC7 front end tube with the filament voltage supplied from the 6.3V PT winding, like all the other tubes. There is no DC applied to the 6SC7 filament from the C model 6L6GB/5881 cathode circuit as there is to the 12SC7 in the B, but the output tube cathodes are still connected to a +20-or-30 B+ voltage through the cathode feedback winding center tap. The B schematic shows +31V on the output tube cathodes. In the B, apparently the 12SC7 filament drops the +31V from the output cathode circuit to around +20 V. From what I measure so far, the CT voltage for the C is around +20-25 Volts DC.
 
Last edited:
I never heard back from Heyboer, so decided with the owner to go for a Hammond OPT replacement without, of course, a cathode feedback winding (more to come on that).

So, I thought I'd start with a description of how the 2200C is different from the 2200B (for which we do have a schematic, as per above). I have attached admittedly crude, but hopefully legible schematics of the major circuit differences.

1. The C model uses at 6SC7 phono pre-amp tube instead of a 12SC7. Further, the 6SC7 filament is powered by the 6.3 VAC filament winding like all the other tubes. A review of the B model 12SC7 shows that it is powered by DC from the bias supply, dropping the nominal 31VDC supply voltage down to about 20VDC, which is then applied to the DC balance pot through two 250 ohm 4W resistors to ground through the balance pot wiper and also to the grids of the output tubes through 470K and 1K resistors. In the C model, the 31VDC supply is connected directly to the DC balance pot through 500 ohm 5W resistors in each leg, and thence to the grids of the output tubes through the same 470K and 1 K grid resistors. Use of DC for the pre-amp tube in the B model is probably for noise control, but all the other tubes are elevated to the same 31VDC through a connection through the wiper of the hum balance pot, as are all the tubes in the C model, including the 6SC7. In the C model, the filaments for all the tubes have 6.3VAC riding on around +30VDC.

2. The B+ supply in the C model has no chokes unlike the B model, which has 2. The filtering in the C model string is all R/C and the values of some of the resistors in the string are different. The voltages, though, are about the same from the take off points to the various tubes. The electrolytic caps in the Model C PS are all the same value and working voltage as the B, but the main power resistors are 4K 10W for the screen supply and 15K 20W for the bias supply in the C, versus 2K 10W and 8K 20W in the B model. The C model also has a 50 ohm 5W wire-wound dropping resistor between the 5U4G rectifier and the first electrolytic caps, probably to replace the DC resistance of the chokes removed in the C model. See schematic.

3. The line fuse is 2A slow blow for the C, compared to 1.8A (?) for the B.

4. The AC line on/off switch is ganged with the treble switch in the C model, instead of the 'loudness compensation' switch as in the B. Why the on/off switch is ganged with the Loudness Compensator in the B I can't say. Also, it is a little strange having the on/off on the Treble control, as you have to re-set the treble whenever you turn the amp on.

5. Another quite significant difference is how the loudness contour function is handled between the two models. Both models have a dual ganged volume control (called 'Gain' in the C model) and in the B the first volume pot is tapped with a 15K resistor in series with a 0.02 uF cap to ground and a "Loudness Compensation" switch that bypasses the cap. In the C model, there is quite a complicated variable loudness control network added between the output wiper of the selector switch and the first gain/volume pot (see schematic). The "Loudness" control pot is mounted on the front panel of the C model and marked in -dB. Interestingly, the C model "Loudness" function is negated when the knob is full CW, and in full effect with the control full CCW--it is definitely not a volume pot, even though it is labeled "Loudness!" I found this out when doing some frequency response testing and getting some mighty strange curves with the "Loudness" control centered. The Loudness control pot has two taps and the R/C network shown in the schematic and provides varying levels of bass and treble boost and attenuation depending on where it is set. I'm going to show some scans later on with this control in various positions. I have never seen this type of set up in any other tube amp--anyone else?

6. The C model has a level control (250K pot) in the 'Radio' input to. This is wired between the 'Radio' RCA input jack and the Selector switch. This 'Radio Level' control pot is located on the rear panel in the C model. (schematic)

So, the C model has six control knobs, left to right: Equalizer, Selector (mode), Loudness, Gain, Bass, and Treble (with on/off) on the front. On the back, you have the hum balance control and the Radio input level pot.

Best I can tell, all the other circuitry is the same for the B and C models and the small tube voltages the same. More on the restoration in subsequent posts.
Appreciate any feedback, comments, questions on all this.
Dave
 

Attachments

  • IMG_4688.JPG
    IMG_4688.JPG
    64.7 KB · Views: 76
  • IMG_4683 copy.jpg
    IMG_4683 copy.jpg
    85.1 KB · Views: 89
  • IMG_4684.JPG
    IMG_4684.JPG
    66 KB · Views: 80
Last edited:
The master gain / loudness knob pair in my Knight preamp is kind of similar. I'd have to dig out a schematic for it, but this rings bells.

Guessing the treble power thing might have been a case of not being able to get a power switch to tack on to the loudness or the gain knob? I think Magnavox put power switches on the bass or treble knob in some of their consoles. Never made any sense to me either.
 
Update: In further work on the C model I discovered that the grid resistors for the output tubes for the C model are 270K, vs. 470K for the B.
 
Last edited:
Hi, All. The work on the Bell 2200C amplifiers is done and I thought I would, as promised, pass along some further information as a possible reference to anyone who runs across the "C" model, as there is not much information available on line for this model. To summarize, I replaced the damaged output transformer with a Hammond 1650AH unit wired in pentode (not ultralinear) mode (as in the stock unit) and adjusted the NFB loop for best balance of stability and response.

As per above, the original transformer used a cathode bias winding in the OPT along with NFB from the 16 ohm tap back to the audio driver tube (6SN7) and this produced quite good measured results in the stock unit (see subsequent posts). Without the cathode winding in the Hammond replacement transformer, I raised the NFB level in the loop to 19 dB to balance stability and response but was not quite able to produce the performance of the stock unit, but I don't think the differences would be audible under nearly all operating circumstances and the amp with the Hammond transformer ended up sounding quite good alongside the stock unit.

So, please forgive the 'data dump' below, but I wanted to provide information that might be useful from this work. As always, any corrections, comments, or observations are most welcome.

Dave
 
Here is a general description of the Bell 2200C amplifier:

Bell 2200C Mono Integrated Amplifiers

July 26, 2018

IMG_4688.JPG

Description:

The Bell 2200C is a 20WPC (RMS) integrated monaural push-pull amplifier, of mid-to-late 50’s vintage with a full range of inputs and controls using 6L6 family output tubes (6L6GB and 5881 are specified but 6L6GC tubes may be used). 6SL7 tubes are used for the audio pre-amp stage and the phase inverter and 6SN7GTA is the audio driver tube. The phono pre-amp tube is a 6SC7 double triode. A 5U4G rectifier is specified and the power supply (unlike the 2200B model) uses R/C filtering throughout the B+ supply.

The amplifier uses NFB coupled from the 16 ohm output tap to the input of the driver stage combined with a tertiary cathode feedback winding, all of which provides quite low measured distortion levels up to maximum output power (< 0.5% THD+N measured 20-20kHz at 1W output and below 0.5% up to rated power at 1 kHz, both on a stock-unit).

Four, 8, and 16 ohm output taps are provided with a hum balance potentiometer in the 6.3VAC heater circuit, which is elevated to about +30VDC by the bias supply through the wiper of the hum pot. Overall bias level is fixed with the output tube cathodes measured around +30 VDC and the control grids operating around +2.5VDC, giving around -27.5V control grid bias levels and quiescent plate current draw of a little over 40 mA for each output tube (with used but tested good output tubes). Output tube current draw is balanced with a single DC BALANCE pot located under the chassis, but there is no adjustment for overall bias level.

Line level inputs on the SELECTOR control include TV, RADIO (with a level adjustment pot on the rear of the chassis), and TAPE. Low-level input is amplified through the 6SC7 phono pre-amp circuit and is selected in the PHONO position (with MAG1 and MAG2 inputs for high and low output magnetic cartridges and a CRYSTAL phono cartridge input jack). A low-cut RUMBLE FILTER position is available for PHONO inputs to counter excessive record or turntable rumble. A MICRO microphone input jack is located atop the chassis with a slide switch to select MICRO or PHONO. The microphone input operates through the second stage of the phono pre-amp.

An EQUALIZER selector has settings for “78 RPM,” “COL-LP,” “RIAA-ORTHO,” “FFRR,” and “EUR” equalization networks.

A “HIGH-Z AUX OUTPUT” jack is also located on top of the chassis with output from the 6SL7 pre-amp available (before the tone controls), presumably to allow the 2200C to be used as a pre-amp stage for a separate power amplifier.

There are tone control knobs for TREBLE and BASS. The on/off switch is ganged with the TREBLE control in the 2200C and the tone controls are active. Playing volume is adjusted by the “GAIN” control, distinguished from the “LOUDNESS” control.

A somewhat unusual continuously variable LOUDNESS control is provided on the front panel as well, demarcated in “-dB” and replacing the (more typical) “Loudness” switch found on many integrated amplifiers through the 50’s and 60’s (including the Bell 2200B model). In the zero position (fully clockwise), the control is not engaged, but as the knob is rotated counter-clockwise, a loudness filter network is engaged that progressively boosts bass and treble response compared to mid-frequencies while reducing the output volume level, thus maintaining the ability to reduce playing volume while boosting the low and high ends, as a more typical switch would do with “Loudness” engaged, as these curves show:

2200C SN583 LOUD 0-5-20-30dB 1W 16dB FB.jpg

The curves are for “LOUDNESS” knob settings (working down from the top) of zero, -5, -20, and -30 dB.

The amplifier is quite sensitive, requiring <0.25 mV in the line level inputs to drive to full output power with the “GAIN” control at maximum.

At the time of this writing, no manuals or schematics can be located for the Bell 2200C model. Although a schematic for the B model is available, close examination of the 2200C model has revealed several significant differences between the B and C model:

1. Power Supply—the C model power supply uses no chokes, unlike the B, and the entire C model B+ string is R/C filtered. The first three resistors in the C model string have different values, but the same wattages. Voltages appear to be similar at the key take-off points. The line fuse value is 2.0A slow blow in the C model and 1.8A SB in the B model.

2. Phono Pre-Amp—The C model uses a 6SC7 tube with the heater supplied from the 6.3VAC winding on the power transformer, compared to a 12SC7 in the B model powered by DC voltage generated in the output tube cathode circuit.

3. Loudness Control—as mentioned above, the C model uses a continuously-variable Loudness control, instead of a single-setting switch in the B model. In the C model, the overall volume is controlled by the Gain control.

4. Radio Level Control—the C model uses a 250K level adjustment pot for this line level input located on the back panel.

5. Bias and Grid Resistors—the C model power resistors in the bias circuit (connected to the ends of the DC balance pot) are 500 ohm 5W units compared to 250 ohm 4W units in the B model. The output tube grid resistors are 270K 1/2W, compared to 470K 1/2W in the B model.

6. Phono Pre-Amp Tube Plate Voltage--The plates in the 6SC7 phono pre-amp tube are wired differently in the C model compared to the B. The second plate (pin 5) take off is wired downstream of a 56K ohm resistor for total resistance of 56K + 27 K ohms = 83K ohms between this plate and 100V supply. This results in reduced voltage for the second (pin 5) plate at around 70VDC compared to the first (pin 2) plate at about 90VDC as measured.

7. On/Off and Loudness Contour Switch—The On/Off power switch is ganged with the Treble control in the C model. In the B model, there is a three-position rotary switch that has Power Off/Power on/Loudness Off settings. That is, the Loudness contour switch turns ON with the power and is switched off in the Loudness Off setting.

8. Equalizer Control Settings—The C model settings are: 78 RPM, COL-LP, RIAA-ORTHO, FFRR, and EUR. In the B model, these settings are: AES, LP/NAB, FLAT, FGN, and FFRR. Switch wiring and the configuration of the equalization networks are, of course, different as well.


Dave
 
Last edited:
Here are measurements of the stock unit after rehab work was completed (mainly replacement of B+ supply electrolytic caps and resistors and complete replacement of Astron and Sangamo paper coupler and bypasss capacitors). The measurements were made with used but tested good 6L6GB output tubes (about 80% rated Gm) and 5U4G rectifier:

Frequency Response 2200C Stock Unit (1W RMS)

2200C SN410 Freq Resp 1W.jpg

Total Harmonic Distortion + Noise (THD+N) vs. Frequency 2200C Stock 1W Output

2200C SN410 THD+N v Freq 1W.jpg

THD+N 15W Output 2200C Stock Unit (1.25 dB below rated power)

2200C SN410 THD+N v. Frequency 15W.jpg

THD+N vs. Power Output 1 kHz 2200C Stock Unit

2200C SN410 THD+N v Power.jpg

Dave
 
Last edited:
Here is before/after rehab view of the chassis underneath (Stock unit with original output transformer):

BEFORE:

IMG_4618.JPG

AFTER (Left Side)

16410 wiring L.jpg

AFTER (Middle):

16410 wiring M.jpg

AFTER (Right Side):

16410 wiring R.jpg



Dave
 
Here are measured voltages and other information for both units compared to the B model schematic:

2200C Voltages.jpg

The 583 unit voltages are measured, of course, after replacement of the output transformer. Some of the values on the 2200B schematic were hard to read even on the best on line copy I could find, so I did my best in noting them.

Dave
 
Appreciate all of the info. I have a "B" that's motorboating and a "C" that is playing but has somewhat low volume and distortion. The "B" had a bad OPT that I sent to Heyboer and they rewired it. Picture of the "C" is before I cleaned it up and put the knobs back on. Going to try to make some headway on them over the Holidays. I also have a couple of 2122s that I use for my bedroom system with some Klipsch 5.2s. Pretty good little amps for what they are.

Bell 2200B.jpg Bell 2200C.jpg Bell 2122 - 1.jpg Bell 2122 - 2.jpg
 
I have one of those. Too bad just one and not a pair, however.
These are the voltages recorded on mine if it is any help:
My Value (Schematic Value)

Wall supply= 120vac (117vac)
5881 plates = 405vdc (402vdc) (0.3vac ripple)
5881 screen = 335vdc (330vdc)
Phase inverter plates = 312vdc/327vdc (295vdc/295vdc)
12sc7 plates = 47vdc/44vdc (32vdc/32vdc)

index.php

Screen Shot 2018-12-16 at 11.31.08 PM.png
 
Last edited:
IC-racer, that looks like a "B" with the filter chokes in the power supply. Some very nice examples of the 2200 from you and Pioneer Vir. Love the gold trim around the power and output transformers. I was pretty impressed with the build quality and performance of these. Haven't seen a 2122 yet, but neat look!

Watch for leaky coupling/bypass caps. The ones in the 2200C's I worked on were paper (despite the plastic molding case) and were very leaky. If you're using the phono input, finding a low-microphonic 6SC7 can be a challenge. Jim McShane recommends the 1625 to replace it and those I got from him were way less noisy and microphonic compared to the installed 6SC7's.

Dave
 
Hi Dave - THANK YOU SOOOOOOOOOOOOO much for your detailing of your rebuild of the Bell 2200C. Of course, I have one with VERY low volume output that increases just a little when the volume is turned up. It hasn't been restored and I'm hoping the output transformer isn't fried. I'll try to get it rewound if it is (a challenge as I don't know how to do it). I'm trying to figure out what to do with with the low volume and how to pinpoint the cause prior to rebuilding it. I'm a mostly retired water guy who LOVES old radio and tube equipment, probably a carryover from my dad who was a repair man in the day. I have the old test equipment and a bunch of the old tubes. I also have 2-2122's. The one where I know it will work with a rebuild sounded fantastic in the day. However, it sounds MUCH better with a weak 6SL7 than a perfect NOS. Don't really understand why as I'm just starting to try and learn how to troubleshoot these. Not gonna lie, I'm fooled quite easily at this point. Just finished a radio (almost fried the power transformer). Good voltages yet minimal sound output that would just go to normal and back to almost nothing. Tubes were good and not shorted. Discovered that one of the old wires was brittle and must've been partially shorted to chassis which couldn't be seen (visually all looked fine). Hope you guys don't mind if I check back if I need help? It may be a couple of months before I can get this going again - trying to move. Thanks again. This is GREAT!!!!
 
Back
Top Bottom