Repairing KRK RP6G2 Rokit 6 Gen2 Speakers

I often browse electronics on craigslist when i need to momentarily 'veg out' at my desk.

Yesterday i came across a pair of KRK Rokit 6 Gen2 speakers. There were a few things about this particular listing that caught my eye:

  • The seller claims one of the speakers has a "capacitor issue" that he cant fix himself
  • The ad claims these speakers are Rokit 6 Gen3. the screws on the front panel and the text on the back are a clear giveaway that they are in-fact Gen2. 
I wasn't really looking to buy speakers, but if the issue is a "capacitor issue"... This looks like a good deal. And it beats only using headphones at my desk all day (although i love my headphones). 

I had a nice pair of Dynaudio BM5A's for the desk that i used with a Benchmark DAC1. But i sold both the Dynaudios, and the Benchmark last year... These KRK's aren't likely as good as the Dynaudios, but sometimes a good deal now is better than nothing. And in this case, if i can buy them broken, fix them, use them for a few months, and sell them for close to break-even... Score!

I did a bit of digging, it turns out there is a common "capacitor issue" with this model speaker. A quick google search uncovered lots of blog posts and youtube videos about it.


In the article above, the author guesses that the capacitor failed because it was pushed beyond its voltage rating. I think that this explanation is possible, but there are other possibilities that are just as likely since the transformer has only a 21.5VAC, and a 15.2VAC RMS tap. 


To calculate DC Voltage from the 21.5VAC-0-21.5VAC tap (which drives the woofer amplifier) through a full wave bridge rectifier:

Tout = 21.5VAC + 21.5VAC = 43VAC
(Tout*1.414) = VACmax = 60.802
Vmax * .637 = Vdc = ~38.7VDC


It is always best to choose a capacitor with voltage rating >= 1.5*voltage
38.7v*1.5=~58v

The stock capacitor is rated at 50V, ideally, 63V would have been better, but 30% overhead should be sufficient (besides, I couldn't find any 3300uF capacitors with 7.5mm lead spacing and 63V rating). I think a couple of equally plausible explanations are that the bridge rectifier ripple output exceeds the ripple current rating of the capacitors. Or a resistor failed, which lead to the capacitor failure. 

This would be a good (upgraded) replacement for the stock 50v smoothing capacitors on the 21.5VAC tap:

In my (yep, I bought it) speaker there happens to be a cracked 2.2kOhm resistor next to one of the smoothing capacitors. And in the article, the author replaced the same 2.2kOhm resistor because it was burnt. 

This is a really handy tool for identifying resistors:

These would be a good (upgraded) replacement for the stock resistors (still staying with 1/2 watt on both amplifiers):

To support the theory of the problem being caused by high ripple current from the rectifier. The noise that comes through the speaker as soon as you turn it on is REALLY LOUD (like full volume loud) and FAST (like 100hz fast). And given that AC from your wall is ~50hz (the frequency fluorescent lights hum at), and the frequency of a full wave rectifier is 2*ACfreq...  

Ohms Law is a thing (you may need this to calculate ideal values of some parts):

Of course, all this can be checked really easily with an oscilloscope tied to the output of the rectifier. But i don't have my oscilloscope handy right now. 

I wont bore you with the formula for approximating ripple voltage, but here is a handy calculator:

My guess is that the amplifier board has a sensitivity of either 1v or 2v (which means it will reach 100% of the amplifiers rated output with a 1v or 2v input signal. So there is a good chance that the ripple is >= 1VAC. To lazy to actually test right now.  

I have a feeling i can fit a capacitor with 10mm lead spacing and a 20mm diameter on the board for the woofer power supply smoothing capacitors, it actually looks like the board is drilled at both 7.5mm and 10mm, and height is not an issue. With those physical constraints, i will be able to find a capacitor with a 63v rating AND significantly higher ripple current rating. Sometimes a brute force solution is the best solution. 

It looks like the rectifier on the 21.5VAC tap is a KBL406G. I suspect they are fine, but here is the part on digikey just in case someone needs them:

The rectifier on the 15.2VAC tap uses 4 discrete diodes. And uses the same 3300uF 50V smoothing capacitors as on the 21.5VAC tap. This must be a newer version of the amp, since many of the youtube videos about repairing this speaker call for 1000uF 35v rated parts here. Given that Vmax in this case is 27.4VAC, capacitors rated at >= 41V are completely adequate. 

The value of smoothing capacitors is generally not critical, and can be as low as 1000uF, and as high as there is physical space for. Bigger is generally better in the case of smoothing capacitors. 

Many of the repair videos and tutorials refer to these large electrolytic capacitors as coupling or filter capacitors. They are not. Coupling or filter capacitors are used to filter undesirable frequencies from the audio signal. The capacitors in questions are referred to as decoupling or smoothing capacitors, and work by both extending the time of the ripple and shorting undesirable AC frequencies to ground. Smoothing capacitors also act as power reserves helping smooth out voltage spikes or rapid demand for power. 


Its worth noting that most of the stock parts in these speakers are not what i would consider to be of "good" quality for a high-resolution monitor. The EI core transformer is noisy, carbon film resistors used throughout are noisy, the rectifiers are noisy, the 4580D opamps are noisy, and i would prefer to see metalized PP capacitors in the audio path rather than small electrolytic and ceramic types. 

At some point, if i keep these speakers long enough, it would like to replace the resistors with vishay-dale metal film type, replace the coupling capacitors with good quality metalized pp parts. and roll opamps (there are 6 4580D in there). 

This thread is a good place to start looking at alternative opamps:

These speakers could be a fun place to do a bit of opamp rolling. The only downside is that the cost of good dip-8 sockets can be a bit high. and these speakers would each use up to 6 of them. 

The woofer "amplifier" seems to be a TDA7296 35v 60w DMOS
This thing produces something like 10% THD at 60w! Did i read that spec right? (THD is around 0.5% at 30W, so it seems these speakers are best used <= half volume)

The volume control attenuates prior to the amplifier stage, so half volume does not mean the volume control in the middle position. It means adjusting the input signal to 50% of the input sensitivity of the amplifier. 

The tweeter "amplifier" is a TDA2052 25v 40w DMOS

Now that we know the tweeter and woofer run from different amplifiers, which have different power supplies (all on the same board). The next test is to see if the loud noise comes out of the woofer, tweeter, or both. 

If we are dealing with the same "capacitor" problem in the article referenced above. The noise should only come out of the woofer. Of course it could also be related to the tweeter amplifier, in which case the noise would only come out of the tweeter. Or the same problem could exist in both amplifiers.

Another issue i noticed is that the insulators over the wire mounted quick connects for the tweeter were pulled back. Its possible there was a short at some point. I hope the DTA2052 is OK. They are cheep and easy to replace, but i'd hate to have to put in another Digikey order (the shipping cost is more than the cost of that part). 

When flipping the power switch on and off quickly, there is a popping sound from the tweeter. Consider putting a 1000pF x2 rated capacitor across the switch to minimize this. This is often referred to as a "spark killer". (upon inspection, i see there is already one there - the sound is probably due to the AC leakage "capacitor issue")

Now that i have considered the theoretical reasons this could loud noise could be happening. Time to test a few things:

I disconnected the tweeter, turned on the amp... and it was silent. Reconnected the tweeter, and there was the noise.

Continued testing is probably a bit of a waste at this point. While the capacitors don't show physical signs of damage, there are very likely shot. Time to replace some capacitors and maybe some resistors.

Parts ordered from Digikey.

As a quick recap, here is what we are going to replace in each speaker (so they match):
  • 2x Smoothing Capacitors for Woofer Amplifier (Digikey 493-11030-ND)
  • 2x 2.2kOhm Resistors for Woofer Amplifier (Digikey CMF2.20KHFCT-ND)
  • 2x Smoothing Capacitors for Tweeter Amplifier (Digikey 493-4614-ND)
  • 2x 2.0Ohm Resistors for Tweeter Amplifier (Digikey CMF2.00KHFCT-ND)
All of the parts ordered are of what i would consider "good" quality for their intended use. They are also significantly uprated from the parts that were previously in place to ensure improved longevity. 

Total cost including shipping, tax, and spare parts (enough to do a 3rd speaker) is ~$66. Or $22 per speaker. 

Its also worth adding that when i got these speakers the cabinets were in pretty rough shape, dirty, splattered with paint, sticky. 

Cleaning and reconditioning products used:
Each of these products were specifically chosen. Formula 409 Multi-Surface Cleaner cannot be substituted for other versions of the 409 product (they won't work properly for cleaning the vinyl coating on the speaker box). 

After cleaning and conditioning the speakers actually look pretty good. There are of course still quite a few cosmetic flaws, they are far from perfect. But given their intended use as a rugged professional monitor. The cosmetics flaws actually add a bit of character... Im tempted to throw a bunch of stickers for some of my favorite brands on the cabinet to further amplify the character quotient. 

From a financial perspective, this project was a fail. The speakers cost $120, the parts cost $66, and the cleaning and conditioning supplies takes us to right around $200 for the pair. With that said, once the parts arrive from Digikey, the speakers will functionally be better than new. And they should have many more years of use in them. 


Update

I finally got around to working on the speakers this weekend. First, i replaced the capacitors and damaged resistors in the broken speaker, plugged it in and got no sound out of the tweeter. The voice coil was destroyed. I swapped in the tweeter from the good speaker, and things worked well... Compared to the speaker that came "good' the repaired one now sounded better. Good parts will do that...

I proceeded to upgrade the capacitors and a few resistors in the other speaker. All good. I have a replacement tweeter on order, with a little luck it will show up this week and i will have a fully functioning, better sounding than new pair of KRK Rockit G2's.

Update 2 

 Replacement tweeter showed up. Speakers assembled, tested... and sounding better than the G3 pair i compared them to.

Update 3

After using the speakers for a while i added snubbers to the bridge rectifier... I think i'm done with modifying these speakers now, i'm pretty happy with how they turned out. Not bad for a total investment of ~$250 + some time. 


Comments

Post a Comment

Popular posts from this blog

PS Audio Directstream, DS MKII, TSS : New DAC or Mods?

PS Audio has been making the Perfectwave digital to audio converter (DAC) for quite some time, in fact, it is now at the tail end of its third incarnation, all using the same casework.  It started with the Perfectwave DAC back in xxxx, in 2012 it was upgraded to the Perfectwave MKII, and then, with the introduction of Ted Smith to the PS Audio Team, the Perfectwave Directstream (aka DS) followed in 2014.  The DS has remained PS Audio’s flagship DAC since 2014, roughly 7 years, with all upgrades being free and applied as firmware updates. Even more impressive is that many audio journalists and audiophiles still consider the DS to be among the best sounding, if not the best measuring, DAC’s on the market.  Full disclosure, the PS Audio Directstream DAC is my personal reference. From what I gather, the last planned firmware update, Sunshine (named for sunshine peak in Colorado) for the DS MKI is scheduled to arrive by the end of April 2021… Though, according to a representative of PS Audi
Read more

PureAudio Lotus DAC5 Review

Image
Work In Progress Review: Background: There isn't much info on this DAC out there yet, but i expect that to change pretty quickly. Here is a foreign language thread i was able to track down: https://translate.google.com/translate?hl=en&sl=zh-TW&u=http://bbs.audiohall.net/viewtopic.php%3Ft%3D11656&prev=search When researching this DAC, I wasn't all that impressed by this unit containing a single ES9038PRO as opposed to some other units that use one for each channel, nor was i impressed by the claimed SNR of > 110db (which represents approximately 18.3 bits of resolution - 110/6), even the ifi micro idsd black claims a SNR > 117db (19.5 bits of resolution - 117/6)... But the fact that the opamps appear to be inserted into sockets for easy rolling is quite interesting, it appears this unit was designed from the ground up for the "tweak" enthusiast. As a benchmark, i typically look for claimed SNR >= 114db (19 bits), and prefer to see clai
Read more