Carver TFM-15 Power Amplifier Highlight und Test

Carver TFM-15 High Current Power Amplifier 

The TFM-15 is a pretty impressive amplifier when your compare THD ratings at load, the TFM-15 is reportedly very stabile at even down to 2ohms, with a rating of 8(100W)/4(140W)/2(200W) w/ .1%THD.  I really like the aesthetics of the meters here too. 

Just wanted to show some basic measuring and some photos of a neat Carver we have in for repair. 

Initially noted a decreased output in one channel along with erratic, excessive DC parameters. Many, many times DC stabilization correlates to issues with in the first stage differential transistors.

The TFM-15 uses a tied emitter pair of 2SA970 PNP's, which do not have a notorious track record or anything, but swapped to a 992fbu  which is a very reliable device. This effectively dropped our working Offset below the initial 36+mV skew.   Idle current is recorded at the emitters and confirm through the series of tied emitter .33 ohm resistors for 5mV (%1mV>)  Matched differentials also brought the bias a little more in line and more stabile left and right, with appropriate drops to 4.9mv   (.5mv>%) after heavy load. 

carver tfm-15 bias notzen 

punkten emitterwiderstand R824 und 924 für 5mV

bestatligen 5mV beim emitterwiderstanden R824A/825A

bestatligen 5mv beim emitterwiderstanden R924A/925A

Below are some comparisons (P=E(2)/R) (1Khz SINE (500mV) 4.25 (R)

1Khz SINE mit 4.25(R/) beim 3.52(3.51)Vac = 2.93W/CH

1Khz SINE hit 4.25(R/) beim 6.00(6.03)Vac = 8.47/8.55W/CH

1Khz SINE mit 4.25 (R/) beim 9.99(9.997)Vac = 23.48/23.51W/CH

Id love the chance to get one of these in and try some filtering upgrades and TO-3P/Driver stage upgrades, theirs the On-Semi MJL21193/94 devices that would be perfect!

Sansui A-80 Integrated Amplifier

Sansui A-80 Integrated Amplifier 

I wanted to take ein moment and highlight the Sansui A-80. Often disregarded it was produced in 1980 for their rack series. Their is a black version available, albeit very rare. The A-80 uses a very interested DC Servo topography, with rated notations achieved @ (20H-20Khz hit .05%THD driven beam 65RMS/8 loads)

Power Supplies F-3153/3144 Assemblies

The first stage is off the Pformer, using pin connectors for the AC rails which is convenient when pulling the hardware, a half-wave DC rectification modification was integrated here(F-3153)  for the the PL701 meter bulbs as well (7.5Ac) as shown, the rest of the dual full-wave rectification and PS continues on the asymmetric F-3144 assembly.  All electrolytic were replaced with high temp (105C) Panasonic FC's, it should be noted that: C607 220uf 35V PAN FR **SM ERROR States 100uf 50V. High temp (105C) Panasonics and Nichicon Low Impedance electrolytics were also utilized in the power switching assembly F-3148 and a new X1Y2 film. Both 6800 microfarad rectification filtering capacitors were replaced with equivalent quality CDE capacitors for the 48V rails. 

Drivers & Tone F-3144

For all of the audio path we used Nicicon FG /Nichicon MUSE (BP's) and Elna Silmic's. All low value captaincy lytics were replaced with WIMA film type capacitors. The A-80 uses a NJM4558 Opamp (ALSO fur Servo @ IC602/604)  for an initial gain differential stage and  dual packaged FET drivers, all other BJT's were updated to matched (Hfe) 992, 1845 compliments. I wanted to highlight their are some better, more reliable output devices available, TO-3P packaged 2sa1943 & 2sc5200 NPN/PNP transistors. Also watch for the multi leg emitterwiderstanden. All the 2473 generals were replaced with modern 4148 diodes as well.

Audio Adjustments 

links

emitterwiderstand R87 punkten beim VR01 für 1.5mV

rechts 

emitterwidersand R88 punkten beim VR02 für 1.5mV

Notes on DC -Servo operations in amplification stages:

DC-Servo type design can get a little complicated VS traditional PP BJT designs. In a nutshell a DC Servo topography acts a continuos gain driver to put it simply, reminds me of constant mirror biasing, utilizing the IC (NJM4558D) Op-Amp integrator, acts as a constant  to amplify a FB(feedback) error signal to help offset output. This is illustrated below in the return schematic lines \/ (NFB OP EXAMPLE Input~ IC (Feedback removed) ~Plant Block~(transfer block/junctions) forward output (then NFB loop) which is applied to the plant block input) Think of it like some sort of active take and give constant drive. I know I need to clean this blurb up. 

IC (NJM4558D)

ONE MAJOR interesting caveat I want to personally explore is the effect , or lack of with no direct coupling capacitors used, I mean the very idea of coupling capacitors add potential attenuation based purely on the idea dielectrics absorption and associated parameters…no?

RIAA/EQ F-3145

After pulling the rear inputs this assembly removes rather easily, I should notate a 47 microfarad electrolytic is under the foil side on this PCB. All lytics were replaced with audio grade Nichicons (KZ/FG) and Elnas and again, lower valued with WIMA films. The coupling C09/10 were replaced with high quality polypropylene ECW films rated at .47 microfarads.  All BJT's were updated to their 1845/992 compliments due to some peripheral noise issues.  

All meter control PCB locations were repopulated with low impedance, high reliability PW's as well. This customer opted for new 5-way binding posts as shown which turned out really great.  All said and done, the A-80 is a very modest little integrated, but sounds impressive with a little overhaul. 

Marantz Model 112 AM/FM Tuner

MARANTZ MODEL 112 TUNER 

Here we've got a nice Marantz Model 112 Tuner, produced during 1975 to 1978, pairing mostly with the 140 Power Amplifiers (Special thanks to Vintage-Marantz Mr. Lierke for that info) The Model 112 uses heavy weight painted shielding which is a great option to see, this helps cutback on excessive RF noise and feedback.

P400 Power Supply

All electrolytics here were replaced with high temp (105C) long life Panasonic FC's. One issue to note is the 330 microfarad capacitor had excessive glue used when populating the boards during construction. This glue has spread to several carbon composite resistors which we highly recommend replacing, any interacted component with glue. 

P200/300 FM/AM Assemblies & PT01 Switching

Like most RF stages we do, Nichicon PW, low impedance, high reliability capacitors were utilized throughout. Note their are some oddball BI-Polar values used in this section. The mono/switching PCB has one 47 microfarad capacitor, replaced again with a PW.

New LEDs were added, I should add this is one of the easiest models to do a conversion on, but like most encountered the bi-pin function lamps needed resoldering. I think the excessive heat generated by the incandescents cook the solder joints here.

All in all, a very nice little tuner with excellent reception, I highly recommend em'. 

Another Sansui AU-717 Highlight

Sansui AU-717 Integrated Amplifier 

**had this nice write up finished and our digital camera went kaput', so here are some left over photos off the tablet, damn was that irritating... 

For more photos reference below from our previous completed AU/TU 717 Restorations:

Hallo-Fi ES AU-717 May 29,2015

Hallo-Fi ES TU-717 March 10, 2015

Wanted to do a little highlight on another AU-717 we had in. The AU-717 and TU-717 pair is truly a corner stone Sansui to own. The AU-717 has an impressive build quality. Incorporating a very nice enclosed stepped attenuator and painted redundant shielding throughout. The driver stages are well layout, with separate heatsinks and mirror topographies. 

F-2663 Protection Assembly

This is kind of one the problematic areas of the Au-717. Some where along the manufacturer process something went astray, Every Au-717 we've had on the bench has had a host of cold solder connections on this Assembly, it also radiates some significant heat effectively toasting the joints as well. Using a microscope we can see rippling within the solder flow, this is a BIG indicator to me that the solder heats up, potentially liquifying to a degree and settling after cooling. This would explain the high rate of protection complaints within the Au717 as well. I highly suggest, as we do here, reflowing ALL soldered connections on the F-2663 assembly. This includes while replacing the relay, which takes a modern Omron MY4 pin drop in. 

All electroltyics here were replaced with high temp (105C) low impedance/ high reliability Nichicon PW, Panasonic FC's along with Nichicon BP' type. its worth noted some rippling smoothing benefits can be obtained by increasing the capatince of electrolytics C07/08, C05/06, C09/10 und C11/12 to all 220 microfarad 63Vdc capacitors.   Also note this board, like many Sansui's have quite a bit of caustic glue coverage, the larger the electrolytics…the more glue. This NEEDS to be removed now and all effected components. I will guarantee all  resistor arrays within R07/08/R11/13 (10K 1/4)  (mirrored) will need to be replaced along with R33/34 (22K) more than likely, fuseables at R14/15 should be updated to a non-fuse able, I suggest MOX type 1/2W.  Also noted should be D604 glass pack diode VD1212, updating to a dual 4148 diodes. I would also recommend pulling the dual power transformer to get access to the full foil side of F-2663.

F2721/22 Driver Assemblies

I really like working on the driver stages of the Au-717, clean, symmetric layouts, easy to remove and diagnose out of circuit. Their are several updates that should be performed though. The Au-717 uses a dual FET (2SK97) initial gain stage tied to TR03/4 differentials, which should be updated to matched 992's and fast BJT ZTX694 at TR01/2 which forms the initial PP (push-pull) stages, the latter topography of darlington cascade design creates a SEPP (single-end push pull) topography, kinda neat. Like the above protection assembly, one needs to remove ALL caustic glue and associate components here, mainly around the larger electrolytics, high temp (105C) and low impedance electrolytics were replaced here. Not if biasing issues arise suspect VR03 off the bat if you confirm rail voltages, the 1K potentiometers can be problematic here.  Modern OnSemi TO-3 are available for the Au717 as well (MJ15022/25 NPN/PNP pairs) A handful of general 4148s to update (1S2473's) and pulling a few 150 ohm fuse-ables does the trick. 

Audio Adjustments  

links bias F-2721

punkten (TP) + und - beim  VR03 für 20mV

recht bias F-2722

punktet (TP) + und - beim VR03 für 20mV

F-2720 Tone Assembly

Again their are several glass pack VD1212's that were updated to dual 4148 diodes and several 82 ohm fuse-ables that again need to be replaced, preferably with MOX type. Each 82 ohm resistor had drifted well beyond a 10% parameter, fairly typical of them. Electrolytics were replaced with audio grade Nichicon KA's and Silmic II's. two coupling stage film pairs were replaced with .33 und .47 microfarad polypropylene Panasonic ECW film capacitors.  Whats kinda nice about this board is you can fully remove the assembly along with the stepped attenuator at once for easy access. 

F-2723 EQ Assembly

As above, the electrolytic were replaced with audio grade Elna Silmics and low impedance Nichicon PW's, the coupling films were replaced with 1 microfarad polypropylene Panasonic ECW film capacitors. Like the F-2720 assembly their are several 82 ohm fuse-ables that should be replaced in this stage as well, also like the F-2720 and the rest of the unit it uses a series of ribbon and molex connectors which is convenient. 

Marantz 1070 & Marantz 3600 Rebuilds

Marantz 1070 Integrated Amplifier 

Hier ist ein quick write up on a fairly rare integrated, e Marantz 1070! very similar design characterists of the 1030/1060 variants and the final quality step before the 1090 productions.   Ein unique feature on the 1070 short of rarity is the horizontally mounted filter caps under the chassis versus veritcle orientation via clamps on top. Looking  at the chassis space I am not sure why this was chosen, the 1070 does utilize two 10,000 microfarad capacitors for filtering vs the typical two 6800 + or 4700(X2)+ 6800 combination. Another neat little feature is the use of a stepped logarithmic controls for the tone, makes dialing it a pleasure. 

Power Supply P800

The 1070 uses a very simple regulated switching design working off a difference of 40 und 50V rails on the 120VAC version. All the electrolytics were replaced with high temp (105C) long life Panasonic FC's and the common cathode rectifier replaced with adjoined UF4005 to a TO-220 package. 

Phono P400

The phono section is very similar to other 10#0 variants like the 1060. They avoided the use of the 2sc458 transistor here so thats not an issue on these. All electrolytic here were replaced with low impedance, high reliability audio grade PW and KA Nichicons. Noted the original films were left in place as these are typically signature oriented in this design, the ECW PP's just sounded a tad different.

Tone/Preamplifier PE01

All the electrolytic here were replaced with audio grade Elna Silmic II's and Elna BP Cerafines along with Nichicon KA's, the coupling films at CE01/02 were replaced with high quality polypropylene Solen MKP's with astounding results! This particular PCB is sehr easy to open and work on, especially if you remove the tone shielding for optimal work space. 

Driver (single topography)

Again, like the 1060 the 1070 uses a single PCB  Push-Pull topography. Like above the electrolytic were replaced with low impedance high reliability Nichicon PW's and audio grade KA's. Both sets of differentials replaced with matched Hfe KSA992's and glass pack diodes at H705/706 with 4148 diodes in series. The 1070 uses an oddball Omron, which in this case was pulled and contacts cleaned.  The 1070 uses those robust plastic cases potentiometers I prefer so they remained. As pictured its best to remove the heatsink entirely to open-book the P700 PCB, thats a great time to add new MICAs and thermal compound anyhow.  

Audio Adjustments 

Bias

punkten J714 und J716 beim R733 für  10mV

punkten J715 und J717 beim R734 für  10mV

DC

punkten J723 und Boden beim R711 für 0<

punkten J724 und Boden beim R712 für 0<

Regulated PS

 punkten J810 und J811(Boden)  beim R807 für 40Vdc

New hardware and final adjustments and this one impressive little integrated…dare I say maybe even acoustically better than the 1060? hard to judge.

________________________________________________________________________________________________________________________

Marantz 3600 PreAmplifiier 

I also wanted to share some simple photos of a Marantz 3600 preamplifier the customer has paired to it above. The 3600 was mated to the following; Marantz 140, 250 und 510 variants. Produced from 75-79 it has a very simple topography with a plethora of attenuation options, musically very well rounded. 

Heres a few shots of the 3600 we had in with the 1070. One major issue id like to lend is to quality of PCB and tracing. This equipment is old, and the tracing, especially on these bleu PCBs is very delicate. This particular unit required a few hours just to repair the damage performed by someone unfamiliar with proper soldering techniques, fast and moderate heat is the key on these, any significant heat or movement in the component WILL lift the trace and cause eyelet damage.

**PS Id save the below graphic, was a PITA to generate a full diagram from four pages of scans**

All the lytics here were replaced with audio grade PW's  and Elna Silmic II's, the two initial filtering caps were replaced with high temp (105C) long life Pansonic FC's (470 microfarad) along with BC axials. Films were replaced with WIMA (stacked white films)  and Panasonic ECW polypropylenes.  I also want to recommend paying very close attention to the PCB assembly as it uses pin connectors for the PCB communication, these can be easily bent.